A multi-source data fusion coal mining machine high-precision positioning system and method

Abstract

This invention discloses a high-precision positioning system and method for a coal mining machine based on multi-source data fusion. The positioning system includes: an optical fiber shape sensor configured to acquire bending shape information of a scraper conveyor during operation; an optical fiber demodulator configured to receive the bending shape information and demodulate it into an electrical signal; an inertial measurement unit configured to acquire the attitude, speed, and position information of the coal mining machine during operation; a shaft encoder configured to acquire linear velocity and linear displacement increment information of the coal mining machine running along the scraper conveyor; a database server configured to receive and store the bending shape information, the attitude, speed, and position information, and the linear velocity and linear displacement increment information; and a computer configured to receive the bending shape information, the attitude, speed, and position information, and the linear velocity and linear displacement increment information, and perform multi-source data fusion correction to achieve high-precision positioning of the coal mining machine.

Description

Technical Field

[0001] This invention relates to the field of intelligent coal mining, and in particular to a high-precision positioning system and method for coal mining machines that integrates multi-source data. Background Technology

[0002] In the process of intelligent coal mine construction moving towards less manned and unmanned operations, the accuracy of collaborative sensing of the working face equipment group has become a technical bottleneck restricting routine autonomous operation. The accurate dynamic sensing of the coal mining machine's position, as a core input parameter of the intelligent mining system, directly affects the reliability of technologies such as autonomous cutting and automatic straightening control. Currently, the position sensing of the coal mining machine mainly relies on single sensing methods such as inertial navigation systems, wireless communication networks, and ultrasonic ranging, which have significant environmental adaptability deficiencies and are difficult to meet the requirements of routine operation of intelligent working faces.

[0003] Research on coal mining machine position fusion sensing mainly focuses on multi-sensor data fusion and error compensation, such as combined positioning using strapdown inertial navigation and wireless sensor networks, and differential inertial sensing components and error compensation methods, providing new approaches for high-precision position sensing of coal mining machines. Therefore, this paper proposes a fusion sensing method based on inertial navigation and fiber optic sensing technologies. This method fuses existing coal mining machine inertial navigation system data with fiber optic sensing system data for the straightness of the scraper conveyor using an unscented Kalman filter, and updates the filter measurement equations using a time-series prediction method. This improves the positioning accuracy of the coal mining machine and lays a solid foundation for the routine operation of intelligent coal mining faces. Summary of the Invention

[0004] This solution addresses the problems and needs raised above by proposing a high-precision positioning system and method for coal mining machines based on multi-source data fusion. The system achieves the aforementioned technical objectives and brings about several other technical benefits due to the adoption of the following technical features.

[0005] This invention proposes a high-precision positioning system for coal mining machines based on multi-source data fusion, comprising:

[0006] A fiber optic shape sensor, installed on a scraper conveyor, is configured to acquire information about the bending shape of the scraper conveyor during operation.

[0007] An optical fiber demodulator, coupled to the optical fiber shape sensor, is configured to receive the bending morphology information and demodulate it into an electrical signal;

[0008] An inertial measurement unit, installed on the coal mining machine, is configured to acquire the attitude, speed, and position information of the coal mining machine during operation;

[0009] A shaft encoder, installed on the gear wheel shaft of the coal mining machine, is configured to acquire information on the linear velocity and linear displacement increment of the coal mining machine as it runs along the scraper conveyor.

[0010] The database server is coupled to the fiber optic demodulator, the inertial measurement unit, and the shaft encoder, respectively, and is configured to receive and store the bending morphology information, the attitude, speed and position information, and the linear velocity and linear displacement increment information of the coal mining machine along the scraper conveyor.

[0011] A computer, coupled to the database server, is configured to receive the bending morphology information, the attitude, speed and position information, and the linear velocity and linear displacement increment information of the coal mining machine along the scraper conveyor, and to perform information fusion correction to achieve high-precision perception of the coal mining machine's position.

[0012] In addition, the high-precision positioning system for coal mining machines based on multi-source data fusion according to the present invention may also have the following technical features:

[0013] In one example of the present invention, the computer includes:

[0014] The first measurement module is configured to calculate the coordinate data of the distribution points on the body of the scraper conveyor based on the bending morphology information.

[0015] The second measurement module is configured to calculate the attitude, speed, and position data of the coal mining machine based on the attitude, speed, and position information.

[0016] The third measurement module is configured to calculate the linear velocity and linear displacement increment data of the coal mining machine along the scraper conveyor based on the linear velocity and linear displacement increment information.

[0017] The fusion correction module includes a matching model, a verification model, a mathematical model, an unscented Kalman filter, and a time-series prediction model. The matching model is configured to calculate the position coordinates of the matching point of the coal mining machine on the scraper conveyor based on the position coordinate data calculated by the first measurement module and the linear velocity and linear displacement increment data calculated by the third measurement module. The verification model is configured to verify the confidence level between the matching point data and the predicted position data of the coal mining machine at the previous moment. The mathematical model is configured to receive the processing results of the verification model and perform calculations using the attitude, velocity, and position data calculated by the second measurement module. The unscented Kalman filter is configured to obtain the optimal estimate of the state of the coal mining machine based on the calculation results of the mathematical model. This optimal estimate is output as the corrected state of the coal mining machine and is simultaneously returned to the second measurement module to correct the errors of the inertial measurement elements. The time-series prediction model receives the optimal estimate, predicts the state of the coal mining machine at the next moment, and feeds the predicted value back to the verification model for correction at the next moment, completing the loop.

[0018] In one example of the invention, the computer further includes:

[0019] The visualization module is configured to obtain accurate coal mining machine position data based on the fusion correction module, and to visualize the two-dimensional and three-dimensional curves of the coal mining machine position through visualization algorithms.

[0020] Another objective of this invention is to provide a positioning method for a high-precision positioning system for a coal mining machine based on multi-source data fusion as described above, comprising the following steps:

[0021] S10: Establish a navigation coordinate system within the mining roadway. n X n Y n Z n , where X n Axis, Y n Axis, Z n The axes point to due east, due north, and the celestial direction, respectively; a coordinate system O for the coal mining machine carrier is established with the installation position of the inertial measurement element as the origin. b X b Y b Z b , where X b The axis is the horizontal axis of the coal mining machine, Y b The axis is the longitudinal axis of the coal mining machine, Z b The axis is the vertical axis of the coal mining machine; a coordinate system O is established at the center point of the scraper conveyor head in the coal mining face. g X g Y g Z g This is used to describe the bending morphology of the scraper conveyor and the orientation of the coal mining machine, where Z... g The axis points in the celestial direction, X g The axis points in the direction of the working face advance, Y g The shaft conforms to the right-hand rule, is parallel to the coal face, and points towards the tail of the machine; a coordinate system O is established with the center of the gear wheel shaft as the origin. m X m Y m Z m , where X m The axis conforms to the right-hand rule and is perpendicular to the direction of travel, Y m Along the direction of travel of the coal mining machine, Z m The axis points in the upward direction;

[0022] S20: The computer calculates the position data of the distribution points on the scraper conveyor body based on the bending morphology information obtained by the fiber optic shape sensor and transmits it to the matching model; the computer calculates the attitude, speed, and position data of the coal mining machine during operation based on the acceleration and angular velocity information obtained by the inertial measurement element and transmits it to the mathematical model; the computer calculates the linear velocity and linear displacement increment data of the coal mining machine along the scraper conveyor based on the linear velocity and linear displacement increment information obtained by the shaft encoder and transmits it to the matching model.

[0023] S30: The matching model is invoked to determine the matching point where the coal mining machine runs onto the scraper conveyor based on the calculated linear displacement increment data, and the scraper conveyor position data in the fiber optic shape sensor corresponding to the position of the coal mining machine is obtained by the computer and transmitted to the verification model.

[0024] S40: The computer calls the verification model and determines whether there is data input from the time series prediction model: if not, the matching point data from the matching model is directly transmitted to the mathematical model; if so, the confidence of the matching point data and the predicted coal mining machine position data at the previous moment is checked, and the data with high confidence is transmitted to the mathematical model.

[0025] S50: After receiving the processing results from the verification model and the attitude, velocity, and position data calculated by the second measurement module, the mathematical model performs state update calculations and transmits them to the unscented Kalman filter to perform optimal estimation of the state of the coal mining machine and obtain the optimal estimate value. The optimal estimate value is output as the corrected state of the coal mining machine, and the visualization module is used to visualize the trajectory curve of the coal mining machine. At the same time, it is returned to the second measurement module to correct the error of the inertial measurement element. The optimal estimate value is transmitted to the dynamic time series prediction model and stored in the sliding window.

[0026] S60: The computer calls the dynamic time-series prediction model and determines whether the number of data n in the sliding window meets the requirements. If not, the prediction process is skipped, and no data is transmitted to the verification model for use in the next moment. The sliding window continues to receive data from the Kalman filter. If the requirements are met, the speed and position of the coal mining machine in the next moment are predicted based on the data in the sliding window. The error covariance matrix of the prediction model is calculated, and the predicted value and the error covariance matrix are transmitted to the verification model for updating the measurement equation in the next moment, thus completing the loop.

[0027] In one example of the present invention, step S20 specifically includes the following steps:

[0028] The computer calculates and obtains the set of discrete point coordinates P of the scraper conveyor body. f :

[0029]

[0030] In the formula, p fi Let be the position coordinates of a discrete point on the body of the scraper conveyor in the n-system;

[0031] The computer 60 calculates and acquires the attitude data set of the coal mining machine during operation based on the inertial measurement element 30. Speed ​​data set V s With location data set P s :

[0032]

[0033] In the formula, ψ i θ i γ i The attitude information of the coal mining machine calculated at a certain moment includes the heading angle, pitch angle, and roll angle; v si p represents the speed information of the coal mining machine at a certain time n; si Let be the position coordinates of the coal mining machine at a certain moment in the n-system;

[0034] The computer calculates and obtains the data set V of the linear velocity V of the coal mining machine along the scraper conveyor. o With linear displacement increment data set △S:

[0035]

[0036] In the formula, v oi The linear velocity of the coal mining machine along the scraper conveyor at a certain moment in the n-system; △s i This represents the displacement increment of the coal mining machine along the scraper conveyor running line at a certain moment in the n-system.

[0037] In one example of the present invention, in step S30, the matching model is invoked to determine the matching point where the coal mining machine runs onto the scraper conveyor based on the calculated linear displacement increment data, and the scraper conveyor position data in the fiber optic shape sensor corresponding to the position of the coal mining machine is obtained by the computer and transmitted to the verification model. Specifically, this includes the following steps:

[0038] Based on the initial position of the coal mining machine on the scraper conveyor, a relative position calculation model for the coal mining machine is constructed using linear displacement increment data and the coal mining machine's motion direction vector T.

[0039]

[0040] In the formula, p o(i+1) This is the relative position data of the coal mining machine;

[0041] After determining the matching point position of the coal mining machine on the scraper conveyor, the computer adopts a differentiated matching strategy to obtain the position p of the coal mining machine.o(i+1) The corresponding scraper conveyor position data in the fiber optic shape sensor includes:

[0042] When the coal mining machine is at position p o(i+1) When a corresponding data point exists in the fiber optic shape sensor, the data point at that location is directly selected as the coordinate data of the matching point for subsequent fusion correction measurement input;

[0043] When the coal mining machine is at position p o(i+1) When no corresponding data point exists in the fiber optic shape sensor, the coordinate data of the data point is obtained by Euclidean distance calculation and used as the measurement input for subsequent fusion correction.

[0044] In one example of the present invention, data points are obtained through Euclidean distance calculation as coordinate data of matching points for subsequent fusion correction measurement input, specifically including the following steps:

[0045] First, let's consider the position p of the coal mining machine. o(i+1) The vertical projection distance d to the reference plane of the scraper conveyor is used as the reference value, and the dynamic error range is set. Then, in the fiber optic shape sensor, a search is performed to find the expression that satisfies p. o(i+1) Euclidean distance d between i Less than (d+) Candidate data points are selected based on the given conditions, and finally, through precision screening, the single data point closest to the baseline value d is selected as the measurement input for subsequent fusion correction.

[0046] In one example of the present invention, in step S40, the confidence level between the matching point data and the predicted coal mining machine position data at the previous moment is checked, and the data with high confidence is transmitted to the mathematical model. This specifically includes the following steps:

[0047] S41: Trajectory data for the next time step based on the filtered coal mining machine speed and position data. Make a prediction and calculate its difference from the actual observed value Y. k+1 Mahalanobis distance D between 2 The expression is as follows:

[0048]

[0049] In the formula, S is the covariance matrix of the observed values;

[0050] S42: Based on the significance level α and the observed value Y k The chi-square right-hand critical value is determined by consulting the chi-square distribution table for the degree of freedom of chi-square m. ;like < Keep Y k+1Perform the unscented Kalman filter estimation at the next time step; if ≥ Then use Replace Y k+1 Using the error covariance matrix R of the prediction model pre The measurement noise matrix R is replaced to optimize and update the unscented Kalman filter measurement equation and perform the filter estimation for the next time step; where the error covariance matrix R of the prediction model is... pre The expression is as follows:

[0051]

[0052] In the formula, e i This represents the prediction error; This represents the mean error.

[0053] In one example of the present invention, in step S41, a dynamic time-series prediction model is invoked to predict the trajectory data for the next time step based on the filtered coal mining machine speed and position data. Making predictions includes:

[0054] Using the position and velocity data from the previous w time steps, the motion information of the coal mining machine in the next time step is analyzed. The prediction is expressed as follows:

[0055]

[0056] In the formula, Here, w represents a predefined time-series prediction model based on the fusion of convolutional neural networks and bidirectional long short-term memory neural networks; p is the length of the sliding time window. si For position; v si For speed;

[0057] The covariance matrix S of the observed values ij The covariance of the observations is obtained, which is used to solve the optimal estimate of the unscented Kalman filter at the next time step, where the covariance matrix S of the observations is... ij The expression is as follows:

[0058]

[0059] In the formula, μ i μ j y(i) and y(j) are the mean values ​​of the observed values ​​Y(i) and Y(j), respectively.

[0060] In one example of the present invention, in step S50, after the mathematical model receives the processing result from the verification model and the attitude, velocity, and position data calculated by the second measurement module, it performs a state update calculation, including the following steps:

[0061] The mathematical model error state vector of the fusion positioning system encompasses attitude error. Velocity error δv, position error δp, gyroscope constant drift error ε b Accelerometer constant drift error▽ b Scale factor error δK D Installation deviation angle error α θ α ψ Its state equation is expressed as follows:

[0062]

[0063] In the formula, x is the state vector; F is the state transition matrix; G is the noise distribution matrix; w is the noise matrix; the expression for the state transition matrix F is as follows:

[0064]

[0065] In the formula, The angular velocity in the n-system; Let be the transformation matrix from the b-system to the n-system;

[0066]

[0067]

[0068]

[0069] In the formula, λ, L, and h represent the longitude, latitude, and altitude of the coal mining machine in the Earth coordinate system, respectively; R M R N These are the radii of the meridian circle and the radii of the east-west circle, respectively.

[0070] Since there is a certain distance between the mounting plane of the inertial measurement element and the plane of the scraper conveyor, in order to calculate the position error of the coal mining machine, it is necessary to use the scraper conveyor position data p obtained in step S30. fi Position data is obtained by projecting it onto the mounting plane of the inertial measurement element. The fusion positioning system is based on the location data. The linear velocity data v oi The position data p calculated by the inertial measurement element are respectively compared with the position data p. si Speed ​​data v si Perform difference calculations and construct multi-source measurement equations:

[0071]

[0072] In the formula, z is the measured value; H is the measurement matrix; and R is the measurement noise matrix.

[0073]

[0074] In the formula, Let be the transformation matrix from the m-system to the b-system; This refers to the linear velocity data in the encoder's own coordinate system.

[0075]

[0076] In the formula, .

[0077] Compared with the prior art, the present invention has the following beneficial effects:

[0078] 1. This solution addresses the pain point of lacking a stable calibration system for inertial measurement elements in underground coal mines. It introduces a fiber optic shape sensing system for scraper conveyors to form a stable combined system. At the same time, it adds a dynamic time-series prediction model to form a stable high-precision positioning system for coal mining machines based on multi-source data fusion, which significantly improves the stability and reliability of the coal mining machine positioning system.

[0079] 2. This method obtains the relative position of the coal mining machine on the scraper conveyor through a shaft encoder independent of the inertial measurement element, and constructs a matching model to achieve the matching of the coal mining machine position to the scraper conveyor shape curve, providing independent measurement information for the coal mining machine positioning system to calibrate the error of the inertial measurement element.

[0080] 3. By introducing a dynamic time series model to obtain the time series characteristics during the operation of the coal mining machine in real time, the limitations of the traditional Kalman filter algorithm in terms of time steps are made up for, which can significantly improve the accuracy of the inertial navigation system in the positioning of the coal mining machine.

[0081] The preferred embodiments of the invention will be described in more detail below with reference to the accompanying drawings, so as to facilitate an understanding of the features and advantages of the invention. Attached Figure Description

[0082] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. The drawings are merely illustrative of some embodiments of the present invention and are not intended to limit the scope of the present invention to all embodiments.

[0083] Figure 1 This is a schematic diagram of a high-precision positioning system for a coal mining machine based on multi-source data fusion according to an embodiment of the present invention.

[0084] Figure 2 This is a schematic diagram of the fusion positioning and sensing of a coal mining machine according to an embodiment of the present invention;

[0085] Figure 3 This is a coal mining machine position matching diagram based on grating measurement points and interpolation data points according to an embodiment of the present invention;

[0086] Figure 4 This is a coal mining machine position matching map based on nearest neighbor dynamic optimization according to an embodiment of the present invention;

[0087] Figure 5 This is a flowchart of the data perception and fusion correction architecture according to an embodiment of the present invention.

[0088] List of reference numerals in the attached diagram:

[0089] Fusion Sensing System 100;

[0090] Fiber optic shape sensor 10;

[0091] Grating measurement point 11;

[0092] Fiber optic demodulator 20;

[0093] Inertial measurement element 30;

[0094] 40 shaft encoder;

[0095] Database server 50;

[0096] Computer 60;

[0097] First measurement module 61;

[0098] Second measurement module 62;

[0099] Third measurement module 63;

[0100] Fusion correction module 64;

[0101] Visualization module 65. Detailed Implementation

[0102] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. The same reference numerals in the drawings represent the same components. It should be noted that the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0103] Unless otherwise defined, the technical or scientific terms used herein shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms “first,” “second,” and similar terms used in this patent application specification and claims do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, “an” or “a” and similar terms do not necessarily indicate a quantity limitation. Terms such as “comprising” or “including” mean that the element or object preceding the word encompasses the element or object listed following the word and its equivalents, without excluding other elements or objects. Terms such as “connected” or “linked” are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as “upper,” “lower,” “left,” and “right” are used only to indicate relative positional relationships; these relative positional relationships may change accordingly when the absolute position of the described object changes.

[0104] According to a first aspect of the present invention, a high-precision positioning system for a coal mining machine based on multi-source data fusion, such as... Figure 1 As shown, it includes:

[0105] An optical fiber shape sensor 10 is mounted on a scraper conveyor and configured to acquire information about the bending shape of the scraper conveyor during operation. For example, the optical fiber shape sensor 10 includes optical fiber measuring points 11 spaced along the scraper conveyor and optical fibers 12 connecting the various optical fiber measuring points.

[0106] The fiber optic demodulator 20 is coupled to the fiber optic shape sensor 10 and is configured to receive the bending shape information and demodulate it into an electrical signal.

[0107] An inertial measurement element 30 is installed on the coal mining machine and configured to acquire the attitude, speed and position information of the coal mining machine during operation.

[0108] A shaft encoder 40 is installed on the gear wheel shaft of the coal mining machine and is configured to acquire information on the linear velocity and linear displacement increment of the coal mining machine as it runs along the scraper conveyor.

[0109] Database server 50 is coupled to the fiber optic demodulator 20, the inertial measurement element 30 and the shaft encoder 40 respectively, and is configured to receive and store the bending shape information, the attitude, speed and position information, and the linear speed and linear displacement increment information of the coal mining machine along the scraper conveyor.

[0110] Computer 60, coupled to database server 50, is configured to receive the bending morphology information, the attitude, speed and position information, and the linear speed and linear displacement increment information of the coal mining machine along the scraper conveyor, and perform information fusion correction to achieve high-precision perception of the coal mining machine's position.

[0111] This positioning system addresses the pain point of lacking a stable calibration system for inertial measurement elements in underground coal mines. It introduces a fiber optic shape sensing system for scraper conveyors to form a stable combined system. At the same time, it adds a dynamic time-series prediction model to form a stable high-precision positioning system for coal mining machines that integrates multi-source data, significantly improving the stability and reliability of the coal mining machine positioning system.

[0112] The positioning system obtains the relative position of the coal mining machine on the scraper conveyor through a shaft encoder independent of the inertial measurement element, and constructs a matching model to achieve the matching of the coal mining machine position to the scraper conveyor shape curve, providing independent measurement information for the coal mining machine positioning system to calibrate the error of the inertial measurement element.

[0113] This positioning system introduces a dynamic time series model to obtain the temporal characteristics of the coal mining machine in real time, which makes up for the limitations of the traditional Kalman filter algorithm in terms of time steps and can significantly improve the accuracy of the inertial navigation system in coal mining machine positioning.

[0114] In one example of the present invention, the computer 60 includes:

[0115] The first measurement module 61 is configured to calculate the coordinate data of the distribution points on the body of the scraper conveyor based on the bending morphology information.

[0116] The second measurement module 62 is configured to calculate the attitude, speed, and position data of the coal mining machine based on the attitude, speed, and position information; and to perform error compensation based on the correction data returned by the fusion correction module 64 to achieve feedback correction.

[0117] The third measurement module 63 is configured to calculate the linear velocity and linear displacement increment data of the coal mining machine along the scraper conveyor based on the linear velocity and linear displacement increment information.

[0118] The fusion correction module 64 includes a matching model, a verification model, a mathematical model, an unscented Kalman filter, and a time-series prediction model. The matching model is configured to calculate the position coordinates of the matching point of the coal mining machine on the scraper conveyor based on the position coordinates calculated by the first measurement module 61 and the linear velocity and linear displacement increment data calculated by the third measurement module 63. The verification model is configured to verify the confidence level between the matching point data and the predicted position data of the coal mining machine at the previous moment. The mathematical model is configured to receive the processing results of the verification model and perform calculations using the attitude, velocity, and position data calculated by the second measurement module 62. The unscented Kalman filter is configured to obtain the optimal estimate of the state of the coal mining machine based on the calculation results of the mathematical model. This optimal estimate is output as the corrected state of the coal mining machine and simultaneously returned to the second measurement module 62 to correct the error of the inertial measurement element 30. The time-series prediction model receives the optimal estimate, predicts the state of the coal mining machine at the next moment, and feeds the predicted value back to the verification model for correction at the next moment, completing the loop.

[0119] In one example of the present invention, the computer 60 further includes:

[0120] The visualization module 65 is configured to obtain accurate coal mining machine position data based on the fusion correction module 64, and to visualize the two-dimensional and three-dimensional curves of the coal mining machine position through a visualization algorithm.

[0121] According to a second aspect of the present invention, a positioning method for a high-precision positioning system for a coal mining machine based on multi-source data fusion as described above is provided. Figure 5 As shown, it includes the following steps:

[0122] S10: Establish a navigation coordinate system within the mining roadway. n X n Y n Z n (n-system), where X n Axis, Y n Axis, Z n The axes point to due east, due north, and the celestial direction, respectively; a coordinate system O for the coal mining machine carrier is established with the installation position of the inertial measurement element 30 as the origin. b X b Y b Z b (b series), where X b The axis is the horizontal axis of the coal mining machine, Y b The axis is the longitudinal axis of the coal mining machine, Z b The axis is the vertical axis of the coal mining machine; a coordinate system O is established at the center point of the scraper conveyor head in the coal mining face. g X g Y g Z g(g series), used to describe the bending shape of the scraper conveyor and the orientation of the coal mining machine, where Z g The axis points in the celestial direction, X g The axis points in the direction of the working face advance, Y g The shaft conforms to the right-hand rule, is parallel to the coal face, and points towards the tail of the machine; a coordinate system O is established with the center of the gear wheel shaft as the origin. m X m Y m Z m (m series), where X m The axis conforms to the right-hand rule and is perpendicular to the direction of travel, Y m Along the direction of travel of the coal mining machine, Z m The axis points upwards; the transformation relationship between coordinate systems is to first rotate and then translate.

[0123] S20: The computer 60 calculates the position data of the distribution points on the scraper conveyor body based on the bending morphology information obtained by the fiber optic shape sensor 10 and transmits it to the matching model; the computer 60 calculates the attitude, speed and position data of the coal mining machine during operation based on the acceleration and angular velocity information obtained by the inertial measurement element 30 and transmits it to the mathematical model; the computer 60 calculates the linear velocity and linear displacement increment data of the coal mining machine along the scraper conveyor based on the linear velocity and linear displacement increment information obtained by the shaft encoder 40 and transmits it to the matching model.

[0124] S30: The matching model is called to determine the matching point where the coal mining machine runs on the scraper conveyor based on the calculated linear displacement increment data, and the scraper conveyor position data in the fiber optic shape sensor 10 corresponding to the position of the coal mining machine is obtained based on the computer 60 and transmitted to the verification model.

[0125] S40: The computer 60 calls the verification model and determines whether there is data input from the time series prediction model: if not, the matching point data from the matching model is directly transmitted to the mathematical model; if so, the confidence of the matching point data and the predicted coal mining machine position data at the previous moment is checked, and the data with high confidence is transmitted to the mathematical model.

[0126] S50: After receiving the processing results from the verification model and the attitude, velocity, and position data calculated by the second measurement module 62, the mathematical model performs state update calculations and transmits them to the unscented Kalman filter to perform optimal estimation of the state of the coal mining machine and obtain the optimal estimate value. The optimal estimate value is output as the corrected state of the coal mining machine and is visualized through the visualization module 65 to display the trajectory curve of the coal mining machine. At the same time, it is returned to the second measurement module 62 to correct the error of the inertial measurement element 30. The optimal estimate value is transmitted to the dynamic time-series prediction model and stored in the sliding window.

[0127] S60: The computer 60 calls the dynamic time-series prediction model and determines whether the number of data points n in the sliding window meets the requirements. If not, the prediction process is skipped, and no data is transmitted to the verification model for use in the next moment. The sliding window continues to receive data from the Kalman filter. If the requirements are met, the speed and position of the coal mining machine in the next moment are predicted based on the data in the sliding window. The error covariance matrix of the prediction model is calculated, and the predicted value and the error covariance matrix are transmitted to the verification model for updating the measurement equation in the next moment, completing the loop. For example, if n reaches 20, if it does, the oldest data will be discarded and the newest data will be accepted when new data enters next time.

[0128] This positioning method addresses the pain point of lacking a stable calibration system for inertial measurement elements in underground coal mines. It introduces a fiber optic shape sensing system for scraper conveyors to form a stable combined system. At the same time, it adds a dynamic time-series prediction model to form a stable high-precision positioning system for coal mining machines that integrates multi-source data, significantly improving the stability and reliability of the coal mining machine positioning system.

[0129] This positioning method obtains the relative position of the coal mining machine on the scraper conveyor through a shaft encoder independent of the inertial measurement element, and constructs a matching model to achieve the matching of the coal mining machine position to the scraper conveyor shape curve, providing independent measurement information for the coal mining machine positioning system to calibrate the error of the inertial measurement element.

[0130] This positioning method introduces a dynamic time series model to obtain the temporal characteristics of the coal mining machine in real time, which makes up for the limitations of the traditional Kalman filter algorithm in terms of time steps and can significantly improve the accuracy of the inertial navigation system in coal mining machine positioning.

[0131] In one example of the present invention, step S20 specifically includes the following steps:

[0132] The computer 60 calculates and obtains the set of discrete point coordinates P of the scraper conveyor body. f :

[0133]

[0134] In the formula, p fi Let be the position coordinates of a discrete point on the body of the scraper conveyor in the n-system;

[0135] The computer 60 calculates and acquires the attitude data set of the coal mining machine during operation based on the inertial measurement element 30. Speed ​​data set V s With location data set P s :

[0136]

[0137] In the formula, ψ i θ i γ i The attitude information of the coal mining machine calculated at a certain moment includes the heading angle, pitch angle, and roll angle; v si p represents the speed information of the coal mining machine at a certain time n; si Let be the position coordinates of the coal mining machine at a certain moment in the n-system;

[0138] The computer 60 calculates and obtains the data set V of the linear velocity V of the coal mining machine along the scraper conveyor. o With linear displacement increment data set △S:

[0139]

[0140] In the formula, v oi The linear velocity of the coal mining machine along the scraper conveyor at a certain moment in the n-system; △s i This represents the displacement increment of the coal mining machine along the scraper conveyor running line at a certain moment in the n-system.

[0141] In one example of the present invention, in step S30, the matching model is invoked to determine the matching point where the coal mining machine runs onto the scraper conveyor based on the calculated linear displacement increment data, and the scraper conveyor position data in the fiber optic shape sensor 10 corresponding to the position of the coal mining machine is obtained by the computer 60 and transmitted to the verification model. Specifically, the steps include the following:

[0142] Based on the initial position of the coal mining machine on the scraper conveyor, a relative position calculation model for the coal mining machine is constructed using linear displacement increment data and the coal mining machine's motion direction vector T.

[0143]

[0144] In the formula, p o(i+1) This is the relative position data of the coal mining machine;

[0145] After determining the matching point position of the coal mining machine on the scraper conveyor, the computer 60 adopts a differentiated matching strategy to obtain the position p of the coal mining machine. o(i+1) The corresponding scraper conveyor position data in the fiber optic shape sensor 10 includes:

[0146] like Figure 3 As shown, when the coal mining machine is at position p o(i+1) When a corresponding data point exists in the fiber optic shape sensor 10, the data point at that location is directly selected as the coordinate data of the matching point for subsequent fusion correction measurement input;

[0147] like Figure 4 As shown, when the coal mining machine is at position po(i+1) When there is no corresponding data point in the fiber optic shape sensor 10, the data point is obtained by Euclidean distance calculation as the coordinate data of the matching point for subsequent fusion correction measurement input.

[0148] In one example of the present invention, data points are obtained through Euclidean distance calculation as coordinate data of matching points for subsequent fusion correction measurement input, specifically including the following steps:

[0149] First, let's consider the position p of the coal mining machine. o(i+1) The vertical projection distance d to the reference plane of the scraper conveyor is used as the reference value, and the dynamic error range is set. Then, in the fiber optic shape sensor, a search is performed to find the expression that satisfies p. o(i+1) Euclidean distance d between i Less than (d+) Candidate data points are selected based on the given conditions. Finally, through precision filtering, the single data point closest to the baseline value d is selected as the measurement input for subsequent fusion and correction. The Euclidean distance calculation formula is as follows:

[0150]

[0151] In the formula, These are the relative position coordinates of the coal mining machine; These are the coordinates of the data points in the fiber optic shape sensor.

[0152] In one example of the present invention, in step S40, the confidence level between the matching point data and the predicted coal mining machine position data at the previous moment is checked, and the data with high confidence is transmitted to the mathematical model. This specifically includes the following steps:

[0153] S41: Trajectory data for the next time step based on the filtered coal mining machine speed and position data. Make a prediction and calculate its difference from the actual observed value Y. k+1 Mahalanobis distance D between 2 The expression is as follows:

[0154]

[0155] In the formula, S is the covariance matrix of the observed values;

[0156] S42: Based on the significance level α and the observed value Y k The chi-square right-hand critical value is determined by consulting the chi-square distribution table for the degree of freedom of chi-square m. ;like < Keep Y k+1 Perform the unscented Kalman filter estimation at the next time step; if ≥ Then use Replace Y k+1 Using the error covariance matrix R of the prediction model pre The measurement noise matrix R is replaced to optimize and update the unscented Kalman filter measurement equation and perform the filter estimation for the next time step; where the error covariance matrix R of the prediction model is... pre The expression is as follows:

[0157]

[0158] In the formula, e i This represents the prediction error; This represents the mean error.

[0159] In one example of the present invention, in step S41, a dynamic time-series prediction model is invoked to predict the trajectory data for the next time step based on the filtered coal mining machine speed and position data. Making predictions includes:

[0160] Using the position and velocity data from the previous w time steps, the motion information of the coal mining machine in the next time step is analyzed. The prediction is expressed as follows:

[0161]

[0162] In the formula, Here, w represents a predefined time-series prediction model based on the fusion of convolutional neural networks and bidirectional long short-term memory neural networks; p is the length of the sliding time window. si For position; v si For speed;

[0163] The covariance matrix S of the observed values ij The covariance of the observations is obtained, which is used to solve the optimal estimate of the unscented Kalman filter at the next time step, where the covariance matrix S of the observations is... ij The expression is as follows:

[0164]

[0165] In the formula, μ i μ j y(i) and y(j) are the mean values ​​of the observed values ​​Y(i) and Y(j), respectively.

[0166] In one example of the present invention, in step S50, after the mathematical model receives the processing result from the verification model and the attitude, velocity, and position data calculated by the second measurement module 62, it performs a state update calculation, including the following steps:

[0167] The mathematical model error state vector of the fusion positioning system encompasses attitude error. Velocity error δv, position error δp, gyroscope constant drift error ε b Accelerometer constant drift error▽ b Scale factor error δK D Installation deviation angle error α θ α ψ Its state equation is expressed as follows:

[0168]

[0169] In the formula, x is the state vector; F is the state transition matrix; G is the noise distribution matrix; w is the noise matrix; the expression for the state transition matrix F is as follows:

[0170]

[0171] In the formula, The angular velocity in the n-system; Let be the transformation matrix from the b-system to the n-system;

[0172]

[0173]

[0174]

[0175] In the formula, λ, L, and h represent the longitude, latitude, and altitude of the coal mining machine in the Earth coordinate system, respectively; R M R N These are the radii of the meridian circle and the radii of the east-west circle, respectively.

[0176] Since there is a certain distance between the mounting plane of the inertial measurement element 30 and the plane of the scraper conveyor, in order to calculate the position error of the coal mining machine, it is necessary to use the scraper conveyor position data p obtained in step S30. fi Position data is obtained by projecting the image onto the mounting plane of the inertial measurement element 30. The fusion positioning system is based on the location data. The linear velocity data v oi The position data p calculated by the inertial measurement element 40 are respectively compared with the position data p. si Speed ​​data v si Perform difference calculations and construct multi-source measurement equations:

[0177]

[0178] In the formula, z is the measured value; H is the measurement matrix; and R is the measurement noise matrix.

[0179]

[0180] In the formula, Let be the transformation matrix from the m-system to the b-system; This refers to the linear velocity data in the encoder's own coordinate system.

[0181]

[0182] In the formula, .

[0183] The foregoing description, with reference to preferred embodiments, details exemplary implementations of the high-precision positioning system and method for coal mining machines based on multi-source data fusion proposed in this invention. However, those skilled in the art will understand that various modifications and alterations can be made to the above specific embodiments without departing from the concept of this invention, and various combinations can be made to the various technical features and structures proposed in this invention without exceeding the protection scope of this invention, which is determined by the appended claims.

Claims

1. A high-precision positioning method for coal mining machines using multi-source data fusion, characterized in that, Includes the following steps: S10: Establish a navigation coordinate system within the mining roadway. O n X n Y n Z n ,in, X n axis, Y n axis, Z n The axes point to due east, due north, and celestial directions, respectively; a coordinate system for the coal mining machine carrier is established with the installation position of the inertial measurement element (30) as the origin. O b X b Y b Z b ,in, X b The shaft is the horizontal shaft of the coal mining machine. Y b The shaft is the longitudinal axis of the coal mining machine. Z b The axis is the vertical axis of the coal mining machine; a coordinate system is established at the center point of the scraper conveyor head in the coal mining face. O g X g Y g Z g This is used to describe the bending shape of the scraper conveyor and the orientation of the coal mining machine, where... Z g The axis points in the sky direction. X g The axis points in the direction of the working face advance. Y g The shaft conforms to the right-hand rule, is parallel to the coal face, and points towards the tail of the machine; a coordinate system is established with the center of the gear wheel shaft as the origin. O m X m Y m Z m ,in, X m The axle conforms to the right-hand rule and is perpendicular to the direction of travel. Y m The shaft runs along the direction of travel of the coal mining machine. Z m The axis points in the upward direction; S20: The computer (60) calculates the position data of the scraper conveyor body distribution points based on the bending morphology information obtained by the fiber optic shape sensor (10) and transmits it to the matching model; the computer (60) calculates the attitude, speed and position data of the coal mining machine during operation based on the acceleration and angular velocity information obtained by the inertial measurement element (30) and transmits it to the mathematical model; the computer (60) calculates the linear velocity and linear displacement increment data of the coal mining machine running along the scraper conveyor based on the linear velocity and linear displacement increment information obtained by the shaft encoder (40) and transmits it to the matching model; S30: The matching model is called to determine the matching point where the coal mining machine runs on the scraper conveyor based on the calculated linear displacement increment data, and the scraper conveyor position data in the fiber optic shape sensor (10) corresponding to the position of the coal mining machine is obtained based on the computer (60) and transmitted to the verification model; S40: The computer (60) calls the verification model and determines whether there is data transmitted from the time-series prediction model: if not, the matching point data from the matching model is directly transmitted to the mathematical model; if so, the confidence of the matching point data and the predicted coal mining machine position data at the previous moment is verified, and the data with high confidence is transmitted to the mathematical model; wherein, the error state vector of the mathematical model covers attitude error, velocity error, position error, gyroscope constant drift error, accelerometer constant drift error, scale factor error and installation deviation angle error; S50: After receiving the processing results from the verification model and the attitude, velocity, and position data calculated by the second measurement module (62) in step S20, the mathematical model performs state update calculations and transmits them to the unscented Kalman filter to perform optimal estimation of the state of the coal mining machine and obtain the optimal estimate value. The optimal estimate value is used as the corrected state output of the coal mining machine and is visualized through the visualization module (65) to display the trajectory curve of the coal mining machine. At the same time, it returns to the second measurement module (62) to correct the error of the inertial measurement element (30). The optimal estimate value is transmitted to the dynamic time-series prediction model and stored in the sliding window. S60: The computer (60) calls the dynamic time series prediction model to determine the number of data in the sliding window. n Does the requirement meet? If not, skip the prediction process. No data is transmitted to the verification model for use in the next moment. The sliding window continues to receive data from the Kalman filter. If the requirement is met, predict the speed and position of the coal mining machine in the next moment based on the data in the sliding window, calculate the error covariance matrix of the prediction model, and transmit the predicted value and error covariance matrix to the verification model for updating the measurement equation in the next moment, thus completing the loop.

2. The high-precision positioning method for coal mining machines based on multi-source data fusion according to claim 1, characterized in that, Step S20 specifically includes the following steps: The computer (60) calculates and obtains the set of discrete point coordinates of the scraper conveyor body. P f : In the formula, p fi for n The coordinates of a discrete point on the body of the scraper conveyor. The computer (60) calculates and obtains a set of attitude data of the coal mining machine during operation based on the inertial measurement element (30). Speed ​​data set V s Location Data Set P s : In the formula, ψ i , θ i , γ i The attitude information of the coal mining machine calculated at a certain moment includes heading angle, pitch angle, and roll angle. v si for n Speed ​​information of the coal mining machine at a certain moment; p si for n The coordinates of the coal mining machine at a certain moment are recorded. The computer (60) calculates and obtains the set of data on the linear velocity of the coal mining machine along the scraper conveyor. V O With linear displacement increment data set △ S : In the formula, v oi for n The linear velocity of the coal mining machine along the scraper conveyor at a certain moment is recorded. △ s i for n The system contains information on the displacement increment of the coal mining machine along the scraper conveyor running line at a certain moment.

3. The high-precision positioning method for coal mining machines based on multi-source data fusion according to claim 1, characterized in that, In step S30, the matching model is invoked to determine the matching point where the coal mining machine runs onto the scraper conveyor based on the calculated linear displacement increment data. The computer (60) obtains the scraper conveyor position data in the fiber optic shape sensor (10) corresponding to the position of the coal mining machine and transmits it to the verification model. Specifically, the steps are as follows: Based on the initial position of the coal mining machine on the scraper conveyor, using linear displacement increment data and combining the coal mining machine's motion direction vector... T Construct a model for calculating the relative position of the coal mining machine: In the formula, p o(i+1) This is the relative position data of the coal mining machine; After determining the matching point position of the coal mining machine on the scraper conveyor, the computer (60) adopts a differentiated matching strategy to obtain the position of the coal mining machine. p o(i+1) The corresponding scraper conveyor position data in the fiber optic shape sensor (10) includes: When the coal mining machine is located p o(i+1) When there is a corresponding data point in the fiber optic shape sensor (10), the data point at that position is directly selected as the coordinate data of the matching point for subsequent fusion correction measurement input; When the coal mining machine is located p o(i+1) When there is no corresponding data point in the fiber optic shape sensor (10), the data point is obtained by Euclidean distance calculation as the coordinate data of the matching point for subsequent fusion correction measurement input.

4. The high-precision positioning method for coal mining machines based on multi-source data fusion according to claim 3, characterized in that, The coordinates of data points obtained through Euclidean distance calculation are used as the measurement input for subsequent fusion and correction. The specific steps include the following: First, based on the location of the coal mining machine p o(i+1) Vertical projection distance to the reference plane of the scraper conveyor d As a benchmark value, and to set the dynamic error range. Then, in the fiber optic shape sensor, a search is performed to find the condition that satisfies the condition. p o(i+1) Euclidean distance between d i Less than ( d + The candidate coordinates of the matching points are selected based on the conditions, and finally, the points that match the baseline value are selected through precision filtering. d The coordinates of the closest single data point are used as the coordinates of the matching point for subsequent measurement input for fusion correction.

5. The high-precision positioning method for coal mining machines based on multi-source data fusion according to claim 1, characterized in that, In step S40, the confidence level between the matching point data and the predicted coal mining machine position data from the previous moment is checked, and the data with high confidence level is transmitted to the mathematical model. This specifically includes the following steps: S41: Trajectory data for the next time step based on the filtered coal mining machine speed and position data. Make predictions and calculate their correlation with actual observations. Y k+1 Mahalanobis distance D 2 The expression is as follows: In the formula, S The covariance matrix of the observations; S42: Based on the level of authorship α Observed values Y k Degrees of freedom card m The right-hand critical value of the chi-square was determined by consulting the chi-square distribution table. ;like < ,reserve Y k+1 Perform the unscented Kalman filter estimation at the next time step; if ≥ Then use Alternative Y k+1 Using the error covariance matrix of the prediction model R pre Alternative Measurement Noise Matrix R This achieves optimized updates to the unscented Kalman filter measurement equations and performs filter estimation for the next time step; where the error covariance matrix of the prediction model is... R pre The expression is as follows: In the formula, e i This represents the prediction error; This represents the mean error.

6. The high-precision positioning method for coal mining machines based on multi-source data fusion according to claim 5, characterized in that, In step S41, a dynamic time-series prediction model is invoked to predict the trajectory data for the next time step based on the filtered coal mining machine speed and position data. Making predictions includes: Before adoption w The position and velocity data at each time step are used to determine the motion information of the coal mining machine at the next time step. The prediction is expressed as follows: In the formula, This is a predefined time-series prediction model based on the fusion of convolutional neural networks and bidirectional long short-term memory neural networks; w The length of the sliding time window; p si For location; v si For speed; Through the covariance matrix of the observed values S ij The covariance matrix of the observations is obtained and used to solve for the optimal estimate of the unscented Kalman filter at the next time step. S ij The expression is as follows: In the formula, μ i μ j Observed values Y ( i ), Y ( j The mean of ).

7. The high-precision positioning method for coal mining machines based on multi-source data fusion according to claim 1, characterized in that, In step S50, after the mathematical model receives the processing results from the verification model and the attitude, velocity, and position data calculated by the second measurement module (62) in step S20, it performs a state update calculation, including the following steps: The mathematical model error state vector of the fusion positioning system encompasses attitude error. Speed ​​error δv Position error δp Gyroscope constant drift error ε b Accelerometer constant drift error▽ b Scale factor error δK D Installation deviation angle error α θ 、α ψ Its state equation is expressed as follows: In the formula, x For state vectors; F This is the state transition matrix; G Assign a matrix to the noise; w The noise matrix; the state transition matrix. F The expression is as follows: In the formula, for n angular velocity under the system; for b Tie n The transformation matrix of the system; In the formula, L , h These represent the dimensions and height of the coal mining machine in the Earth coordinate system; R M , R N These are the radii of the meridian circle and the radii of the east-west circle, respectively. Since there is a certain distance between the mounting plane of the inertial measurement element (30) and the plane of the scraper conveyor, it is necessary to calculate the position error of the coal mining machine based on the scraper conveyor position data obtained in step S30. p fi Position data is obtained by projecting the image onto the mounting plane of the inertial measurement element (30). The fusion positioning system is based on the location data. The linear velocity data v oi The position data respectively calculated by the inertial measurement element (30) p si Speed ​​data v si Perform difference calculations and construct multi-source measurement equations: In the formula, z These are measured values; H For measurement matrix; R For measuring the noise matrix; In the formula, for m Tie b The transformation matrix of the system; This refers to the linear velocity data in the encoder's own coordinate system. In the formula, .

8. A high-precision positioning system for a coal mining machine based on multi-source data fusion, used to implement the high-precision positioning method for a coal mining machine based on multi-source data fusion as described in any one of claims 1-7, characterized in that, include: An optical fiber shape sensor (10) is installed on the scraper conveyor and configured to acquire information on the bending shape of the scraper conveyor during operation. The fiber demodulator (20) is coupled to the fiber shape sensor (10) and configured to receive the bending shape information and demodulate it into an electrical signal; An inertial measurement element (30) is installed on the coal mining machine and configured to acquire the attitude, speed and position information of the coal mining machine during operation. A shaft encoder (40) is installed on the gear wheel shaft of the coal mining machine and is configured to acquire information on the linear velocity and linear displacement increment of the coal mining machine as it runs along the scraper conveyor. The database server (50) is coupled to the fiber optic demodulator (20), the inertial measurement element (30), and the shaft encoder (40) respectively, and is configured to receive and store the bending morphology information, the attitude, speed and position information, and the linear speed and linear displacement increment information of the coal mining machine along the scraper conveyor. A computer (60), coupled to the database server (50), is configured to receive the bending morphology information, the attitude, speed and position information, and the linear speed and linear displacement increment information of the coal mining machine running along the scraper conveyor, and to perform information fusion correction to achieve high-precision perception of the coal mining machine position.

9. The high-precision positioning system for coal mining machines based on multi-source data fusion according to claim 8, characterized in that, The computer (60) includes: The first measurement module (61) is configured to calculate the coordinate data of the distribution points of the scraper conveyor body based on the bending morphology information. The second measurement module (62) is configured to calculate the attitude, speed and position data of the coal mining machine based on the attitude, speed and position information; The third measurement module (63) is configured to calculate the linear velocity and linear displacement increment data of the coal mining machine along the scraper conveyor based on the linear velocity and linear displacement increment information; The fusion correction module (64) includes a matching model, a verification model, a mathematical model, an unscented Kalman filter, and a time-series prediction model. The matching model is configured to calculate the position coordinate data of the matching point of the coal mining machine on the scraper conveyor based on the position coordinate data calculated by the first measurement module (61) and the linear velocity and linear displacement increment data calculated by the third measurement module (63). The verification model is configured to verify the confidence of the matching point data and the predicted position data of the coal mining machine at the previous moment. The mathematical model is configured to receive the processing results of the verification model and calculate the attitude, velocity, and position data calculated by the second measurement module (62). The unscented Kalman filter is configured to obtain the optimal estimate value by performing the optimal estimation of the state of the coal mining machine through the calculation results of the mathematical model. The optimal estimate value is used as the output of the corrected state of the coal mining machine and is returned to the second measurement module (62) to correct the error of the inertial measurement element (30). The time-series prediction model is configured to receive the optimal estimate value, predict the state of the coal mining machine at the next moment, and feed the predicted value back to the verification model for correction at the next moment, thus completing the loop.

10. The high-precision positioning system for coal mining machines based on multi-source data fusion according to claim 8, characterized in that, The computer (60) also includes: The visualization module (65) is configured to obtain accurate coal mining machine position data based on the fusion correction module (64), and to visualize the two-dimensional and three-dimensional curves of the coal mining machine position through the visualization algorithm.

Images