Mining equipment positioning system and method based on multi-source data

Abstract

The invention discloses a multi-source data-based mining equipment positioning system and method. The system is characterized in that a plurality of ultra-wideband base stations are sequentially mounted in a crossheading roadway at intervals; the coding identifiers are mounted on supporting columns of the hydraulic supports in a one-to-one correspondence manner, and are positioned on one side close to the coal mining machine; the inertial sensing unit, the camera, the ultra wide band positioning module and the controller are all mounted on the body of the coal mining machine; the visual angle of the camera faces a supporting column of the hydraulic support. The ultra-wideband positioning module is positioned on one side, close to a crossheading roadway, of a coal mining machine body and is connected with the ultra-wideband base station; the controller is respectively connected with the inertial sensing unit, the camera and the ultra-wideband base station. The method comprises the following steps: carrying out initial position calibration by using the ultra-wideband positioning module; predicting the position of the coal mining machine based on inertial navigation; updating the prediction state of the inertial sensing unit based on the identification of the coding standard by the camera; and the positioning continuity is ensured by fusing the data of the inertial sensing unit and the camera. According to the system and the method, real-time and accurate positioning operation of the coal mining machine in the working process can be realized.

Description

Technical Field

[0001] The present invention relates to the field of intelligent positioning technology, and in particular to a mining equipment positioning system and method based on multi-source data. Background Art

[0002] In the fully mechanized mining face in underground coal mines, the precise positioning of the coal mining machine is an important basis for ensuring safe production and improving coal mining efficiency. The accuracy and reliability of the positioning method directly determine the automation level of the fully mechanized mining face.

[0003] At present, the positioning of coal mining machines mainly relies on manual observation, wireless radio frequency identification or inertial navigation. Relying on workers to observe the position of the coal mining machine on the working surface, and then combining the length of the working surface and the bracket number to determine the current position of the coal mining machine, this method has large errors and increases the labor intensity of workers. At the same time, it has great safety hazards. When wireless radio frequency identification is applied to coal mines, the signal attenuation is serious, and a large number of radio frequency tags need to be arranged underground, which greatly increases the complexity and maintenance cost of the system. When using inertial navigation technology to locate the coal mining machine, the acceleration and angular velocity of the coal mining machine are mainly measured by the accelerometer and gyroscope in the inertial measurement unit, and then the displacement and attitude change are obtained by integrating these data, thereby determining the position of the coal mining machine. Inertial navigation does not need to rely on external signals and is suitable for complex underground environments. However, inertial navigation has cumulative errors, and long-term operation will cause positioning drift and gradually reduce accuracy. In addition, the strong vibration of the coal mining machine will aggravate the positioning drift, resulting in inaccurate positioning results. In order to effectively solve the limitations of traditional positioning technology, it is urgent to provide a coal mining machine positioning method that is suitable for complex underground environments and high-precision positioning requirements. Summary of the invention

[0004] In view of the problems existing in the above-mentioned prior art, the present invention provides a mining equipment positioning system and method based on multi-source data. The system has a simple structure, low maintenance cost, low operating energy consumption, ideal positioning effect and high positioning accuracy. It can realize real-time and accurate positioning operation of the coal mining machine during operation, ensure the safety and reliability of the coal mining machine during operation, and have good social and economic benefits. The method utilizes unscented Kalman filtering to fuse the data obtained by the inertial sensor unit, camera and ultra-wideband positioning module, which can significantly improve the positioning accuracy and positioning automation level of the coal mining machine during long-term operation. It can reduce the complexity of the positioning system, and at the same time, can effectively reduce the dependence on manual operation and reduce the cost of manpower input.

[0005] In order to achieve the above-mentioned object, the present invention provides a mining equipment positioning system based on multi-source data, including a plurality of hydraulic supports located in a coal mining face and a coal mining machine for cutting coal seams, and also includes an ultra-wideband base station, a coding mark, an inertial sensing unit, a camera, an ultra-wideband positioning module and a controller;

[0006] Multiple ultra-wideband base stations are installed in the chute tunnel in sequence and at intervals, and are used as positioning nodes of the ultra-wideband positioning module;

[0007] The number of the coding marks is the same as the number of the hydraulic supports, and the coding marks are installed on the pillars of the hydraulic supports in a one-to-one correspondence and are located on a side of the pillars close to the coal mining machine;

[0008] The inertial sensor unit, camera, ultra-wideband positioning module and controller are all installed on the body of the coal mining machine; the viewing angle of the camera is perpendicular to the direction of the coal mining working face and faces the pillars of the hydraulic support; the ultra-wideband positioning module is located on the side of the coal mining machine body close to the drift tunnel, and is connected to the ultra-wideband base station through wireless communication;

[0009] The controller is connected to the inertial sensing unit, the camera and the ultra-wideband base station respectively.

[0010] Furthermore, in order to improve the recognition accuracy of the camera, the height of the camera viewing angle is the same as the installation height of the coding mark.

[0011] Furthermore, in order to facilitate the real-time display of the positioning result of the coal mining machine, a display is also included, and the display is connected to the controller and is used to display the positioning result data of the coal mining machine in real time.

[0012] As a preference, the controller is a PLC controller.

[0013] In the present invention, by installing an inertial sensor unit on the upper side of the coal mining machine, the position of the coal mining machine in a moving state can be predicted more accurately, and at the same time, a higher response speed can be ensured during the mobile positioning process. By installing a coding mark on the support of the hydraulic support, and at the same time, installing a camera on the coal mining machine, and making the camera's viewing angle face the hydraulic support, when the coal mining machine moves to the coding mark, the camera can be used to identify the coding mark, and then the position information corresponding to the coding mark can be obtained, so that the position information corresponding to each hydraulic support can be used to calibrate the data predicted by the inertial sensor unit at intervals, which can effectively reduce or eliminate the cumulative error of the predicted state of the inertial sensor unit, and ensure the positioning accuracy of the coal mining machine during continuous operation. By installing multiple ultra-wideband base stations at intervals in the upper and lower sides of the chute roadway, and installing an ultra-wideband positioning module on the side of the coal mining machine close to the chute roadway, the ultra-wideband positioning module can be used to locate the coal mining machine at the end of the working face, and the obtained positioning data can be reliably transmitted to the controller with the ultra-wideband base station as a relay, ensuring the reliable transmission of the positioning data of the ultra-wideband positioning module. In this way, when the coal mining machine returns to the end of the working face, the predicted state of the inertial sensor unit can be initially calibrated using the ultra-wideband positioning module that has better positioning performance for static objects, ensuring that the inertial sensor unit has an accurate reference in the initial state.

[0014] The system has a simple structure, low maintenance cost, low operating energy consumption, ideal positioning effect, and high positioning accuracy. It can realize real-time and accurate positioning of the coal mining machine during operation, ensuring the safety and reliability of the coal mining machine during operation, and has good social and economic benefits.

[0015] The present invention also provides a mining equipment positioning method based on multi-source data, using a mining equipment positioning system based on multi-source data, comprising the following steps:

[0016] Step 1: Arrange a mining equipment positioning system based on multi-source data;

[0017] Multiple ultra-wideband base stations are installed in sequence and at intervals in the drift tunnel;

[0018] An inertial sensor unit, a camera and an ultra-wideband positioning module are respectively installed on the body of the coal mining machine, wherein the viewing angle of the camera is fixed and faces the hydraulic support and is perpendicular to the direction of the coal mining working face, and the ultra-wideband positioning module is installed on the side of the coal mining machine body close to the drift roadway and is connected to the ultra-wideband base station through wireless communication;

[0019] A unique coding mark is installed on the column of each hydraulic support, and multiple position information corresponding to the multiple coding marks is stored in the controller as a position data set; wherein the installation height of the coding mark is consistent with the installation height of the camera, and the coding mark is made of reflective material and covered with an antistatic coating;

[0020] Step 2: Use the ultra-wideband positioning module to perform initial position calibration;

[0021] S21: First, use formula (1) to calibrate the state x of the mining equipment positioning system based on multi-source data;

[0022] x=[p,v,q,b a ,b ω ] T (1);

[0023] Where p is the position vector, which indicates the specific position coordinates of the coal mining machine in three-dimensional space, and is used to determine the precise position of the coal mining machine on the working surface; v is the velocity vector, which indicates the moving speed of the coal mining machine in all directions, and is used to reflect the motion state of the coal mining machine; q is the attitude quaternion, which is used to describe the rotation attitude of the coal mining machine in space. The quaternion form avoids the singularity problem of Euler angle representation; b a is the inertial navigation accelerometer bias, which represents the systematic error of the accelerometer in the inertial navigation system and needs to be estimated and corrected in the filtering process; b ω is the gyroscope bias, which represents the systematic error of the gyroscope and also needs to be estimated and corrected in the filtering process;

[0024] S22: When the coal mining machine is located at the end of the working face, the initial position information of the coal mining machine relative to the reference point is obtained by using the ultra-wideband positioning module (UWB), and is sent to the controller via the ultra-wideband base station. Then, the controller obtains the initial state x0 of the mining equipment positioning system based on multi-source data based on the initial position information and formula (2); in this way, the measurement value of the ultra-wideband positioning module can be used as the initial state of the mining equipment positioning system based on multi-source data;

[0025] x0=[p0,v0,q0,b a,0 ,b ω,0 ] T (2);

[0026] Where p0 is the initial position of the coal mining machine, p0 = z UWB , z UWBis the initial position information measured by the ultra-wideband positioning module; v0 is the initial speed of the coal mining machine at the end of the working face and not yet started, v0 = 0; q0 is the initial posture of the coal mining machine at the end of the working face and not yet started, and the value is 0 or an estimated value set according to the posture state, which can be estimated and corrected by the unscented Kalman filter later; b a,0 is the initial value of the inertial navigation accelerometer bias, which is set to 0 and can be estimated and corrected by the unscented Kalman filter later; b ω,0 is the initial value of the gyroscope bias, which is set to 0 and can be estimated and corrected by the unscented Kalman filter later;

[0027] Step 3: Predict the position of the coal mining machine based on inertial navigation;

[0028] S31: Before the coal mining machine starts working from the end of the working face and runs to the first coding mark in front, the position of the coal mining machine is located by using the inertial sensor unit. During this process, the inertial sensor unit collects the inertial sensor data of the coal mining machine in real time during movement and sends it to the controller. The controller uses the unscented Kalman filter to predict the current state and predicts the acceleration a according to formula (3): k , according to formula (4) the angular velocity ω is predicted k ;

[0029]

[0030]

[0031] In the formula, is the inertial navigation measurement value of the acceleration at the current moment, is the inertial navigation measurement value of the angular velocity at the current moment, b a,k-1 is the deviation term of acceleration, b ω,k-1 is the deviation term of angular velocity;

[0032] S32: The controller predicts the attitude angle q according to formula (5) kk-1 , according to formula (6) the speed v is predicted kk-1 , according to formula (7) the position p is predicted kk-1 ;

[0033]

[0034] v kk-1 =v k-1 +(R(q k ) k +g)Δt(6);

[0035]

[0036] In the formula, δq kk-1 is the angular velocity ω k The calculated tiny quaternion, is the symbol for quaternion multiplication; R(q k ) is the rotation matrix calculated from the attitude quaternion measured by inertial navigation, g is the gravity acceleration vector, and Δt is the time step between adjacent moments;

[0037] S33: Update the covariance P between the predicted state and the measured value of the inertial sensor unit according to formula (8) k ;

[0038]

[0039] In the formula, F k-1 is the state transfer matrix, Q k-1 is the process noise covariance matrix;

[0040] Step 4: Update the predicted state of the inertial sensing unit based on the camera's recognition of the coding standard;

[0041] S41: Use the camera to collect image data in real time, and match the coding mark through the feature recognition algorithm; when the coal mining machine runs to the coding mark, the camera matches the coding mark and extracts the feature point information in the coding mark. At the same time, the camera recognizes the coordinates (u, v) of the coding mark in the camera through the image recognition algorithm, and then sends the feature point information and coordinates (u, v) to the controller; the controller compares the received feature point information with the pre-stored position data set, matches and determines the position information corresponding to the coding mark. At the same time, the controller obtains the measurement matrix h(x) of the camera positioning according to formula (9);

[0042]

[0043] In the formula, f x With f y is the focal length of the camera in pixels, which is obtained directly through camera calibration; c x With c y are the principal point coordinates in pixel units, obtained directly through camera calibration;

[0044] S42: The controller obtains the predicted value z of the camera at time k according to formula (10) cam,k , according to formula (11), the camera measurement residual y is obtained k ; Use the camera's measurement results to correct the predicted value of the inertial sensor unit, and obtain the state x of the mining equipment positioning system based on multi-source data after the camera is updated according to formula (12): k ;

[0045] z cam,k=h(x kk-1 )(10);

[0046]

[0047] x k =x kk-1 +K k y k (12);

[0048] In the formula, is the measurement value of the camera at time k, K k is the Kalman gain at time k, Among them, P xz is the covariance between the camera's estimated state and the measured value, H k is the Jacobian matrix of the measurement matrix h(x) to the state, R k is the measurement noise covariance matrix, S k is the measurement covariance;

[0049] Step 5: Before the coal mining machine leaves the last coding mark and arrives at the next coding mark, execute step 3 to locate the coal mining machine through the predicted value of the inertial sensor unit; when the coal mining machine runs to the next coding mark, execute step 4 to correct the predicted value of the inertial sensor unit through the absolute measurement value of the camera; in this way, execute steps 3 and 4 repeatedly to locate the coal mining machine by fusing the predicted value of the inertial sensor unit and the absolute measurement value of the camera until the coal mining machine returns to the end after the current cutter has finished cutting;

[0050] Step 6: When the coal mining machine returns to the end of the working face, re-execute step 2 and use the ultra-wideband positioning module to calibrate the initial position p0 of the coal mining machine. The remaining parameters in the initial state x0 are not initialized. After initialization, repeat steps 3 to 5 to achieve continuous positioning of the coal mining machine and output the positioning results in real time.

[0051] Furthermore, in order to ensure the reliability of the communication of the ultra-wideband positioning module, in step 1, the distance between adjacent base stations is 10 meters.

[0052] In the present invention, when the coal mining machine starts working from the end of the working face or returns to the end after completing a round of work, on the one hand, the ultra-wideband positioning module can be used for initial setting, and the advantages of high precision, real-time and strong anti-interference ability of the ultra-wideband positioning module can be used to efficiently and accurately realize the initial alignment of the predicted state of the inertial sensing unit; on the other hand, the error accumulated by the inertial sensing unit over a long period of operation can be effectively eliminated, and the accumulation of drift problems can be effectively prevented; at the same time, through this regular initial position update, it can be ensured that the system can maintain high accuracy and reliability throughout the entire working process, avoiding the positioning deviation caused by the long-term operation of inertial navigation, and ensuring the stability and accuracy of the positioning of the coal mining machine. Before the coal mining machine leaves the previous coding mark and arrives at the next coding mark, the position of the coal mining machine is positioned by using the inertial sensing unit, which can give full play to the significant advantages of the inertial sensing unit in terms of high positioning accuracy and high dynamic response for continuous motion trajectories. Since the inertial sensing unit relies on the integral calculation of acceleration and angular velocity, a large position drift will occur over time, while the coding mark position provided by the camera has a higher absolute accuracy. To this end, a coding mark is installed on the support of each hydraulic support, and a camera is installed on the coal mining machine. In this way, every time the coal mining machine runs to the coding mark, the camera can obtain the corresponding position information by identifying the coding mark. Then, every time the camera identifies the coding mark on the hydraulic support, the state estimation is updated by the unscented Kalman filter, so as to use the camera's direct observation results of the coding standard to achieve effective calibration of the predicted value of the inertial sensor unit, which can effectively reduce or eliminate the cumulative error of the predicted state of the inertial sensor unit. In this way, by fusing the data of the inertial sensor unit and the camera, the continuity of positioning can be continued to be guaranteed by relying on the inertial sensor unit when the camera cannot detect the coding mark, and the deviation can be effectively corrected when the camera recognizes the coding mark, so that before the coal mining machine returns to the end, the predicted state of the inertial sensor unit can be calibrated in intervals to eliminate the cumulative error in the prediction process of the inertial sensor unit, so as to give full play to the advantages of the two positioning technologies of the inertial sensor unit and the camera, and achieve a higher precision and robust positioning effect, ensuring that the coal mining machine can still accurately perform positioning operations in complex environments. In the process of fusing the inertial sensor unit and camera data, the unscented Kalman filter method is used. Since the filter update mechanism continuously improves the state estimate through each new measurement data obtained, when the camera captures the position information corresponding to the new coding mark, the filter will use the difference between these measurement data and the previous inertial sensor unit predicted state, that is, the measurement residual, combined with the Kalman gain to update the state. The Kalman gain determines the degree of influence of the measurement data on the state estimate correction. Through this process, the filter not only updates the state estimate, but also updates the covariance matrix at the same time, thereby gradually reducing the uncertainty and error of the state.As time goes by, more and more measurement data will enter the filter, which will make the coal mining machine state estimation more accurate by fusing the prediction of the inertial sensing unit and the absolute measurement of the camera, gradually reducing the positioning error.

[0053] This method implements multi-source data fusion positioning technology on the coal mining machine, and uses the unscented Kalman filter to fuse the data obtained by the inertial sensor unit, camera and ultra-wideband positioning module. It can significantly improve the positioning accuracy and positioning automation level of the coal mining machine during long-term operation, and reduce the complexity of the positioning system. At the same time, it can effectively reduce the dependence on manual operation, reduce the cost of manpower input, and significantly reduce the probability of safety accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a schematic diagram of the arrangement of the positioning system in the working surface of the present invention;

[0055] Figure 2 It is a flow chart of the positioning method in the present invention. DETAILED DESCRIPTION

[0056] The present invention will be further described below.

[0057] like Figure 1 As shown, the present invention provides a mining equipment positioning system based on multi-source data, including a plurality of hydraulic supports located in a coal mining face and a coal mining machine for cutting coal seams, and also includes an ultra-wideband base station, a coding mark, an inertial sensing unit, a camera, an ultra-wideband positioning module and a controller;

[0058] Multiple ultra-wideband base stations are installed in the chute tunnel in sequence and at intervals, and are used as positioning nodes of the ultra-wideband positioning module;

[0059] The number of the coding marks is the same as the number of the hydraulic supports, and the coding marks are installed on the pillars of the hydraulic supports in a one-to-one correspondence and are located on a side of the pillars close to the coal mining machine;

[0060] The inertial sensor unit, camera, ultra-wideband positioning module and controller are all installed on the body of the coal mining machine; the viewing angle of the camera is perpendicular to the direction of the coal mining working face and faces the pillars of the hydraulic support; the ultra-wideband positioning module is located on the side of the coal mining machine body close to the drift tunnel, and is connected to the ultra-wideband base station through wireless communication;

[0061] The controller is connected to the inertial sensing unit, the camera and the ultra-wideband base station respectively.

[0062] In order to improve the recognition accuracy of the camera, the height of the camera viewing angle is the same as the installation height of the coding mark.

[0063] In order to facilitate the real-time display of the positioning results of the coal mining machine, a display is also included, and the display is connected to the controller and is used to display the positioning result data of the coal mining machine in real time.

[0064] As a preference, the controller is a PLC controller.

[0065] In the present invention, by installing an inertial sensor unit on the upper side of the coal mining machine, the position of the coal mining machine in a moving state can be predicted more accurately, and at the same time, a higher response speed can be ensured during the mobile positioning process. By installing a coding mark on the support of the hydraulic support, and at the same time, installing a camera on the coal mining machine, and making the camera's viewing angle face the hydraulic support, when the coal mining machine moves to the coding mark, the camera can be used to identify the coding mark, and then the position information corresponding to the coding mark can be obtained, so that the position information corresponding to each hydraulic support can be used to calibrate the data predicted by the inertial sensor unit at intervals, which can effectively reduce or eliminate the cumulative error of the predicted state of the inertial sensor unit, and ensure the positioning accuracy of the coal mining machine during continuous operation. By installing multiple ultra-wideband base stations at intervals in the upper and lower sides of the chute roadway, and installing an ultra-wideband positioning module on the side of the coal mining machine close to the chute roadway, the ultra-wideband positioning module can be used to locate the coal mining machine at the end of the working face, and the obtained positioning data can be reliably transmitted to the controller with the ultra-wideband base station as a relay, ensuring the reliable transmission of the positioning data of the ultra-wideband positioning module. In this way, when the coal mining machine returns to the end of the working face, the predicted state of the inertial sensor unit can be initially calibrated using the ultra-wideband positioning module that has better positioning performance for static objects, ensuring that the inertial sensor unit has an accurate reference in the initial state.

[0066] The system has a simple structure, low maintenance cost, low operating energy consumption, ideal positioning effect, and high positioning accuracy. It can realize real-time and accurate positioning of the coal mining machine during operation, ensuring the safety and reliability of the coal mining machine during operation, and has good social and economic benefits.

[0067] like Figure 2 As shown, the present invention also provides a mining equipment positioning method based on multi-source data, using a mining equipment positioning system based on multi-source data, including the following steps:

[0068] Step 1: Arrange a mining equipment positioning system based on multi-source data;

[0069] Multiple ultra-wideband base stations are installed in sequence and at intervals in the drift tunnel;

[0070] An inertial sensor unit, a camera and an ultra-wideband positioning module are respectively installed on the body of the coal mining machine, wherein the viewing angle of the camera is fixed and faces the hydraulic support and is perpendicular to the direction of the coal mining working face, and the ultra-wideband positioning module is installed on the side of the coal mining machine body close to the drift roadway and is connected to the ultra-wideband base station through wireless communication;

[0071] A unique coding mark is installed on the column of each hydraulic support, and multiple position information corresponding to the multiple coding marks is stored in the controller as a position data set; wherein the installation height of the coding mark is consistent with the installation height of the camera, and the coding mark is made of reflective material and covered with an antistatic coating;

[0072] Step 2: Use the ultra-wideband positioning module to perform initial position calibration;

[0073] S21: First, use formula (1) to calibrate the state x of the mining equipment positioning system based on multi-source data;

[0074] x=[p,v,q,b a ,b ω ] T (1);

[0075] Where p is the position vector, which indicates the specific position coordinates of the coal mining machine in three-dimensional space, and is used to determine the precise position of the coal mining machine on the working surface; v is the velocity vector, which indicates the moving speed of the coal mining machine in all directions, and is used to reflect the motion state of the coal mining machine; q is the attitude quaternion, which is used to describe the rotation attitude of the coal mining machine in space; b a is the inertial navigation accelerometer bias, which represents the systematic error of the accelerometer in the inertial navigation system; b ω is the gyroscope bias, which represents the systematic error of the gyroscope;

[0076] S22: When the coal mining machine is located at the end of the working face, the initial position information of the coal mining machine relative to the reference point is obtained by using the ultra-wideband positioning module, and sent to the controller via the ultra-wideband base station. Then, the controller obtains the initial state x0 of the mining equipment positioning system based on multi-source data based on the initial position information and formula (2);

[0077] x0=[p0,v0,q0,b a,0 ,b ω,0 ] T (2);

[0078] Where p0 is the initial position of the coal mining machine, p0 = z UWB , z UWBis the initial position information measured by the ultra-wideband positioning module; v0 is the initial speed of the coal mining machine at the end of the working face and not yet started, v0 = 0; q0 is the initial posture of the coal mining machine at the end of the working face and not yet started, and the value is 0 or an estimated value set according to the posture state; b a,0 is the initial value of the inertial navigation accelerometer bias, which is 0; b ω,0 is the initial value of the gyroscope bias, which is 0;

[0079] Step 3: Predict the position of the coal mining machine based on inertial navigation;

[0080] S31: Before the coal mining machine starts working from the end of the working face and runs to the first coding mark in front, the position of the coal mining machine is located by using the inertial sensor unit. During this process, the inertial sensor unit collects the inertial sensor data of the coal mining machine in real time during movement and sends it to the controller. The controller uses the unscented Kalman filter to predict the current state and predicts the acceleration a according to formula (3): k , according to formula (4) the angular velocity ω is predicted k ;

[0081]

[0082]

[0083] In the formula, is the inertial navigation measurement value of the acceleration at the current moment, is the inertial navigation measurement value of the angular velocity at the current moment, b a,k-1 is the deviation term of acceleration, b ω,k-1 is the deviation term of angular velocity;

[0084] S32: The controller predicts the attitude angle q according to formula (5) kk-1 , according to formula (6) the speed v is predicted kk-1 , according to formula (7) the position p is predicted kk-1 ;

[0085]

[0086] v kk-1 =v k-1 +(R(q k ) k +g)Δt(6);

[0087]

[0088] In the formula, δq kk-1 is the angular velocity ω k The calculated tiny quaternion, is the symbol for quaternion multiplication; R(q k ) is the rotation matrix calculated from the attitude quaternion measured by inertial navigation, g is the gravity acceleration vector, and Δt is the time step between adjacent moments;

[0089] S33: Update the covariance P between the predicted state and the measured value of the inertial sensor unit according to formula (8) k ;

[0090]

[0091] In the formula, F k-1 is the state transfer matrix, Q k-1 is the process noise covariance matrix, which can be set according to random white noise or selected from experience before camera correction;

[0092] Step 4: Update the predicted state of the inertial sensing unit based on the camera's recognition of the coding standard;

[0093] S41: Use the camera to collect image data in real time, and match the coding mark through the feature recognition algorithm; when the coal mining machine runs to the coding mark, the camera matches the coding mark and extracts the feature point information in the coding mark. At the same time, the camera recognizes the coordinates (u, v) of the coding mark in the camera through the image recognition algorithm, and then sends the feature point information and coordinates (u, v) to the controller; the controller compares the received feature point information with the pre-stored position data set, matches and determines the position information corresponding to the coding mark. At the same time, the controller obtains the measurement matrix h(x) of the camera positioning according to formula (9). The measurement matrix h(x) is used to connect the pixel coordinates measured by the camera with the state of the system;

[0094]

[0095] In the formula, f x With f y is the focal length of the camera in pixels, which is obtained directly through camera calibration; c x With c y are the principal point coordinates in pixel units, obtained directly through camera calibration;

[0096] S42: The controller obtains the predicted value z of the camera at time k according to formula (10) cam,k , according to formula (11), the camera measurement residual y is obtained k ; Use the camera's measurement results to correct the predicted value of the inertial sensor unit, and obtain the state x of the mining equipment positioning system based on multi-source data after the camera is updated according to formula (12): k ;

[0097] z cam,k =h(xkk-1 )(10);

[0098]

[0099] x k =x kk-1 +K k y k (12);

[0100] In the formula, is the measurement value of the camera at time k, K k is the Kalman gain at time k, Among them, P xz is the covariance between the camera's estimated state and the measured value, H k is the Jacobian matrix of the measurement matrix h(x) to the state, R k is the measurement noise covariance matrix, S k is the measurement covariance;

[0101] Step 5: Before the coal mining machine leaves the last coding mark and arrives at the next coding mark, execute step 3 to locate the coal mining machine through the predicted value of the inertial sensor unit; when the coal mining machine runs to the next coding mark, execute step 4 to correct the predicted value of the inertial sensor unit through the absolute measurement value of the camera; in this way, execute steps 3 and 4 repeatedly to locate the coal mining machine by fusing the predicted value of the inertial sensor unit and the absolute measurement value of the camera until the coal mining machine returns to the end after the current cutter has finished cutting;

[0102] Step 6: When the coal mining machine returns to the end of the working face, re-execute step 2 and use the ultra-wideband positioning module to calibrate the initial position p0 of the coal mining machine. The remaining parameters in the initial state x0 are not initialized. After initialization, repeat steps 3 to 5 to achieve continuous positioning of the coal mining machine and output the positioning results in real time.

[0103] In order to ensure the reliability of the communication of the ultra-wideband positioning module, in step 1, the distance between adjacent base stations is 10 meters.

[0104] In the present invention, when the coal mining machine starts working from the end of the working face or returns to the end after completing a round of work, on the one hand, the ultra-wideband positioning module can be used for initial setting, and the advantages of high precision, real-time and strong anti-interference ability of the ultra-wideband positioning module can be used to efficiently and accurately realize the initial alignment of the predicted state of the inertial sensing unit; on the other hand, the error accumulated by the inertial sensing unit over a long period of operation can be effectively eliminated, and the accumulation of drift problems can be effectively prevented; at the same time, through this regular initial position update, it can be ensured that the system can maintain high accuracy and reliability throughout the entire working process, avoiding the positioning deviation caused by the long-term operation of inertial navigation, and ensuring the stability and accuracy of the positioning of the coal mining machine. Before the coal mining machine leaves the previous coding mark and arrives at the next coding mark, the position of the coal mining machine is positioned by using the inertial sensing unit, which can give full play to the significant advantages of the inertial sensing unit in terms of high positioning accuracy and high dynamic response for continuous motion trajectories. Since the inertial sensing unit relies on the integral calculation of acceleration and angular velocity, a large position drift will occur over time, while the coding mark position provided by the camera has a higher absolute accuracy. To this end, a coding mark is installed on the support of each hydraulic support, and a camera is installed on the coal mining machine. In this way, every time the coal mining machine runs to the coding mark, the camera can obtain the corresponding position information by identifying the coding mark. Then, every time the camera identifies the coding mark on the hydraulic support, the state estimation is updated by the unscented Kalman filter, so as to use the camera's direct observation results of the coding standard to achieve effective calibration of the predicted value of the inertial sensor unit, which can effectively reduce or eliminate the cumulative error of the predicted state of the inertial sensor unit. In this way, by fusing the data of the inertial sensor unit and the camera, the continuity of positioning can be continued to be guaranteed by relying on the inertial sensor unit when the camera cannot detect the coding mark, and the deviation can be effectively corrected when the camera recognizes the coding mark, so that before the coal mining machine returns to the end, the predicted state of the inertial sensor unit can be calibrated in intervals to eliminate the cumulative error in the prediction process of the inertial sensor unit, so as to give full play to the advantages of the two positioning technologies of the inertial sensor unit and the camera, and achieve a higher precision and robust positioning effect, ensuring that the coal mining machine can still accurately perform positioning operations in complex environments. In the process of fusing the inertial sensor unit and camera data, the unscented Kalman filter method is used. Since the filter update mechanism continuously improves the state estimate through each new measurement data obtained, when the camera captures the position information corresponding to the new coding mark, the filter will use the difference between these measurement data and the previous inertial sensor unit predicted state, that is, the measurement residual, combined with the Kalman gain to update the state. The Kalman gain determines the degree of influence of the measurement data on the state estimate correction. Through this process, the filter not only updates the state estimate, but also updates the covariance matrix at the same time, thereby gradually reducing the uncertainty and error of the state.As time goes by, more and more measurement data will enter the filter, which will make the coal mining machine state estimation more accurate by fusing the prediction of the inertial sensing unit and the absolute measurement of the camera, gradually reducing the positioning error.

[0105] This method implements multi-source data fusion positioning technology on the coal mining machine, and uses the unscented Kalman filter to fuse the data obtained by the inertial sensor unit, camera and ultra-wideband positioning module. It can significantly improve the positioning accuracy and positioning automation level of the coal mining machine during long-term operation, and reduce the complexity of the positioning system. At the same time, it can effectively reduce the dependence on manual operation, reduce the cost of manpower input, and significantly reduce the probability of safety accidents.

Claims

1. A mining equipment positioning system based on multi-source data, comprising a plurality of hydraulic supports located in a coal mining working face and a coal mining machine for cutting coal seams, characterized in that: It includes an ultra-wideband base station, a coding mark, an inertial sensing unit, a camera, an ultra-wideband positioning module and a controller; Multiple ultra-wideband base stations are installed in the chute tunnel in sequence and at intervals, and are used as positioning nodes of the ultra-wideband positioning module; The number of the coding marks is the same as the number of the hydraulic supports, and the coding marks are installed on the pillars of the hydraulic supports in a one-to-one correspondence and are located on a side of the pillars close to the coal mining machine; The inertial sensor unit, camera, ultra-wideband positioning module and controller are all installed on the body of the coal mining machine; the viewing angle of the camera is perpendicular to the direction of the coal mining working face and faces the pillars of the hydraulic support; the ultra-wideband positioning module is located on the side of the coal mining machine body close to the drift tunnel, and is connected to the ultra-wideband base station through wireless communication; The controller is connected to the inertial sensing unit, the camera and the ultra-wideband base station respectively.

2. The mining equipment positioning system based on multi-source data according to claim 1, characterized in that: The height of the camera viewing angle is the same as the installation height of the coding mark.

3. The mining equipment positioning system based on multi-source data according to claim 1, characterized in that: It also includes a display, which is connected to the controller and is used to display the positioning result data of the coal mining machine in real time.

4. The mining equipment positioning system based on multi-source data according to claim 3 is characterized in that: The controller is a PLC controller.

5. A mining equipment positioning method based on multi-source data, using a mining equipment positioning system based on multi-source data as claimed in any one of claims 1 to 4, characterized in that: The following steps are involved: Step 1: Arrange a mining equipment positioning system based on multi-source data; Multiple ultra-wideband base stations are installed in sequence and at intervals in the drift tunnel; An inertial sensor unit, a camera and an ultra-wideband positioning module are respectively installed on the body of the coal mining machine, wherein the viewing angle of the camera is fixed and faces the hydraulic support and is perpendicular to the direction of the coal mining working face, and the ultra-wideband positioning module is installed on the side of the coal mining machine body close to the drift roadway and is connected to the ultra-wideband base station through wireless communication; A unique coding mark is installed on the column of each hydraulic support, and multiple position information corresponding to the multiple coding marks is stored in the controller as a position data set; wherein the installation height of the coding mark is consistent with the installation height of the camera, and the coding mark is made of reflective material and covered with an antistatic coating; Step 2: Use the ultra-wideband positioning module to perform initial position calibration; S21: First, use formula (1) to calibrate the state x of the mining equipment positioning system based on multi-source data; x=[p,v,q,b a ,b ω ] T (1); Where p is the position vector, which indicates the specific position coordinates of the coal mining machine in three-dimensional space, and is used to determine the precise position of the coal mining machine on the working surface; v is the velocity vector, which indicates the moving speed of the coal mining machine in all directions, and is used to reflect the motion state of the coal mining machine; q is the attitude quaternion, which is used to describe the rotation attitude of the coal mining machine in space; b a is the inertial navigation accelerometer bias, which represents the systematic error of the accelerometer in the inertial navigation system; b ω is the gyroscope bias, which represents the systematic error of the gyroscope; S22: When the coal mining machine is located at the end of the working face, the initial position information of the coal mining machine relative to the reference point is obtained by using the ultra-wideband positioning module, and sent to the controller via the ultra-wideband base station. Then, the controller obtains the initial state x0 of the mining equipment positioning system based on multi-source data based on the initial position information and formula (2); x0=[p0,v0,q0,b a,0 ,b w,0 ] T (2); Where p0 is the initial position of the coal mining machine, p0 = z UWB , z UWB is the initial position information measured by the ultra-wideband positioning module; v0 is the initial speed of the coal mining machine at the end of the working face and not yet started, v0 = 0; q0 is the initial posture of the coal mining machine at the end of the working face and not yet started, and the value is 0 or an estimated value set according to the posture state; b a,0 is the initial value of the inertial navigation accelerometer bias, which is 0; b ω,0 is the initial value of the gyroscope bias, which is 0; Step 3: Predict the position of the coal mining machine based on inertial navigation; S31: Before the coal mining machine starts working from the end of the working face and runs to the first coding mark in front, the position of the coal mining machine is located by using the inertial sensor unit. During this process, the inertial sensor unit collects the inertial sensor data of the coal mining machine in real time during movement and sends it to the controller. The controller uses the unscented Kalman filter to predict the current state and predicts the acceleration a according to formula (3): k , according to formula (4) the angular velocity ω is predicted k ; In the formula, is the inertial navigation measurement value of the acceleration at the current moment, is the inertial navigation measurement value of the angular velocity at the current moment, b a,k-1 is the deviation term of acceleration, b ω,k-1 is the deviation term of angular velocity; S32: The controller predicts the attitude angle q according to formula (5) kk-1 , according to formula (6) the speed v is predicted kk-1 , according to formula (7) the position p is predicted kk-1 ; v kk-1 =v k-1 +(R(q k )a k +g)Δt(6); In the formula, δq kk-1 is the angular velocity ω k The calculated tiny quaternion, is the symbol for quaternion multiplication; R(q k ) is the rotation matrix calculated from the attitude quaternion measured by inertial navigation, g is the gravity acceleration vector, and Δt is the time step between adjacent moments; S33: Update the covariance P between the predicted state and the measured value of the inertial sensor unit according to formula (8) k ; In the formula, F k-1 is the state transfer matrix, Q k-1 is the process noise covariance matrix; Step 4: Update the predicted state of the inertial sensing unit based on the camera's recognition of the coding standard; S41: Use the camera to collect image data in real time, and match the coding mark through the feature recognition algorithm; when the coal mining machine runs to the coding mark, the camera matches the coding mark and extracts the feature point information in the coding mark. At the same time, the camera recognizes the coordinates (u, v) of the coding mark in the camera through the image recognition algorithm, and then sends the feature point information and coordinates (u, v) to the controller; the controller compares the received feature point information with the pre-stored position data set, matches and determines the position information corresponding to the coding mark. At the same time, the controller obtains the measurement matrix h(x) of the camera positioning according to formula (9); In the formula, f x With f y is the focal length of the camera in pixels, which is obtained directly through camera calibration; c x With c y are the principal point coordinates in pixel units, obtained directly through camera calibration; S42: The controller obtains the predicted value z of the camera at time k according to formula (10) cam,k , according to formula (11), the camera measurement residual y is obtained k ; Use the camera's measurement results to correct the predicted value of the inertial sensor unit, and obtain the state x of the mining equipment positioning system based on multi-source data after the camera is updated according to formula (12): k ; z cam,k =h(x kk-1 )(10); x k =x kk-1 +K k y k (12); In the formula, is the measurement value of the camera at time k, K k is the Kalman gain at time k, Among them, P xz is the covariance between the camera's estimated state and the measured value, H k is the Jacobian matrix of the measurement matrix h(x) to the state, R k is the measurement noise covariance matrix, S k is the measurement covariance; Step 5: Before the coal mining machine leaves the last coding mark and arrives at the next coding mark, execute step 3 to locate the coal mining machine through the predicted value of the inertial sensor unit; when the coal mining machine runs to the next coding mark, execute step 4 to correct the predicted value of the inertial sensor unit through the absolute measurement value of the camera; in this way, execute steps 3 and 4 repeatedly to locate the coal mining machine by fusing the predicted value of the inertial sensor unit and the absolute measurement value of the camera until the coal mining machine returns to the end after the current cutter has finished cutting; Step 6: When the coal mining machine returns to the end of the working face, re-execute step 2 and use the ultra-wideband positioning module to calibrate the initial position p0 of the coal mining machine. The remaining parameters in the initial state x0 are not initialized. After initialization, repeat steps 3 to 5 to achieve continuous positioning of the coal mining machine and output the positioning results in real time.

6. The method for locating mining equipment based on multi-source data according to claim 5, characterized in that: In step 1, the distance between adjacent base stations is 10 meters.

Images