A multi-source information fusion positioning system and method for trackless rubber-tyred vehicles in coal mines

Abstract

The present invention relates to the field of positioning and navigation of auxiliary transportation vehicles in underground coal mines, and specifically to a multi-source information fusion positioning system and method for trackless rubber-tyred vehicles in underground coal mines. The method establishes a wheel odometer kinematic model of the trackless rubber-tyred vehicle by installing an encoder module on the non-steering wheel of the trackless rubber-tyred vehicle. The heading angle solved by the kinematic model and the heading angle solved by the strapdown inertial navigation are fused by using an extended Kalman filter, and a dead reckoning algorithm based on multiple encoders and strapdown inertial navigation is obtained from the wheel mileage information and the fused heading angle, thereby improving the accuracy of dead reckoning. A strapdown inertial navigation / dead reckoning combined positioning model is established based on an improved adaptive extended Kalman filter to achieve secondary fusion of the encoder and the strapdown inertial navigation. A combined positioning fault-tolerant model is constructed to achieve detection and isolation of abnormal driving conditions such as slipping or sliding of the trackless rubber-tyred vehicle. The camera module on the trackless rubber-tyred vehicle and the QR code arranged in the lane are used to achieve staged calibration of the positioning information of the trackless rubber-tyred vehicle.

Description

Technical Field

[0001] The invention belongs to the technical field of positioning of auxiliary transport vehicles in underground coal mines, and in particular relates to a multi-source information fusion positioning system and method for trackless rubber-tyred vehicles in underground coal mines. Background Art

[0002] In the actual production process of coal, auxiliary transportation is one of the most important links. Trackless rubber-tyred vehicles have the advantages of flexibility, strong climbing ability, low operating cost and high transportation efficiency. They are gradually replacing trams and becoming the main auxiliary transportation equipment in coal mines. Since coal mines are completely enclosed spaces, the geological conditions are complex and changeable, the working environment is harsh, and the tunnels are narrow and long with many branches, traditional subjective driving has safety hazards such as traffic jams and collisions, which affect the safe and efficient production of coal. In view of the above problems, it is necessary to establish an intelligent dispatching and management system for trackless rubber-tyred vehicles in coal mines, and realizing real-time and accurate positioning of trackless rubber-tyred vehicles is an important part of it.

[0003] At present, the positioning technologies used in coal mines mainly include wireless positioning technology and inertial navigation technology. Wireless positioning technology includes laser scanning positioning technology, Bluetooth positioning technology, Wi-Fi positioning technology, RFID positioning technology, ZigBee positioning technology and UWB positioning technology. The application of these technologies has promoted the development of positioning technology in coal mines. However, due to problems such as signal attenuation caused by wireless signal reflection, non-line-of-sight scenarios and multipath effects, the application of various wireless positioning technologies in coal mines is subject to certain restrictions. Although inertial navigation technology has strong autonomy and is not affected by the external environment, it has the problem of cumulative errors.

[0004] Considering that various positioning technologies have their own advantages and disadvantages and applicable scenarios, it is difficult to provide a reliable and low-cost positioning solution with a single technology. To address this problem, an effective solution is to improve the accuracy and stability of the entire positioning system by combining positioning technologies with different characteristics. However, the research on the positioning technology of rubber-tyred trackless vehicles in coal mines is still insufficient. There are problems such as poor autonomy, high cost, difficulty in orientation, and insufficient reliability. In addition, the influence of abnormal conditions such as wheel slippage or sliding on positioning accuracy has not been fully considered. In order to solve the positioning problem of rubber-tyred trackless vehicles in complex environments, a positioning method based on the fusion of multi-encoders and inertial navigation is proposed. Combining kinematic modeling, dead reckoning algorithm, strapdown inertial navigation algorithm, adaptive extended Kalman filter algorithm, combined positioning fault-tolerant model and visual two-dimensional code recognition, low-cost, high-precision and reliable positioning of rubber-tyred trackless vehicles in coal mines is achieved, providing a certain theoretical basis for the construction of intelligent auxiliary transportation systems in coal mines. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to provide a multi-source information fusion positioning system and method for trackless rubber-tyred vehicles in coal mines to solve the problem of low-cost and high-precision positioning of trackless rubber-tyred vehicles in coal mines.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme: a multi-source information fusion positioning method for trackless rubber-tyred vehicles in coal mines, characterized in that the method comprises the following steps:

[0007] S1: A kinematic model of a wheel odometer of a rubber-tyred trackless vehicle is established based on the encoder data installed on the non-steering wheel of the rubber-tyred trackless vehicle, and the forward speed, heading angle and heading angular velocity information of the rubber-tyred trackless vehicle are obtained by solving the kinematic model;

[0008] S2: Perform strapdown inertial navigation calculation based on the strapdown inertial navigation module data to obtain the attitude, speed and position information of the rubber-tyred trackless vehicle;

[0009] S3: The extended Kalman filter algorithm is used to fuse the heading data calculated based on the wheel odometer kinematic model and the heading angle calculated by the strapdown inertial navigation system. The dead reckoning algorithm based on the multi-encoder and strapdown inertial navigation system is obtained from the wheel mileage information and the fused heading angle information.

[0010] S4: Establish a strapdown inertial navigation / dead reckoning combined positioning model to achieve secondary fusion of encoder and strapdown inertial navigation;

[0011] S5: Establish a combined positioning fault-tolerant model to detect and isolate abnormal driving conditions of rubber-tyred trackless vehicles and improve the anti-interference ability of the combined positioning system;

[0012] S6: a plurality of QR codes containing position information are set at the vehicle shelter chamber in the underground coal mine tunnel, and the transportation route of the trackless rubber-tyred vehicle is divided into a plurality of driving sections using the position information set by the QR code, and the QR code in the tunnel is scanned and identified using a camera module installed on the trackless rubber-tyred vehicle to achieve staged calibration and resetting of the positioning information of the trackless rubber-tyred vehicle;

[0013] S7: The relative positioning information based on the combination of multiple encoders and strapdown inertial navigation depth, as well as the absolute positioning information obtained by scanning and recognizing the QR code with a monocular camera, are used to realize the combination of local relative positioning and global absolute positioning of the rubber-tyred trackless vehicle in the underground tunnel of the coal mine.

[0014] Furthermore, the step S2 also includes: the strapdown inertial navigation solution algorithm is as follows:

[0015] Quaternion attitude update algorithm:

[0016]

[0017] In the formula, tk The quaternion of the carrier's attitude change at each moment, For n series from t k-1 to k The quaternion of the attitude change at the moment, t k-1 The quaternion of the attitude change of system b relative to system n at the moment, For b series from t k-1 to k Quaternion of attitude change at a given moment.

[0018] Speed ​​update algorithm:

[0019]

[0020] In the formula, and t k-1 and t k The inertial navigation solution speed at the moment, and They are update cycles T k =t k -t k-1 The velocity increment of the specific force velocity increment and the velocity increment of the harmful acceleration in the internal navigation coordinate system.

[0021] Location update algorithm:

[0022]

[0023] In the formula, p k and p k-1 t k and t k-1 The inertial navigation solution position at the time,

[0024]

[0025] Furthermore, the step S3 further comprises: k y k z k θ k γ k ψ k ] T Defined as the state variable of the extended Kalman filter, the posture data calculated by the wheel odometer kinematic model is used as the prediction model to establish the extended Kalman filter state equation, and the posture data calculated by the strapdown inertial navigation system is used as the measurement model to establish the extended Kalman filter measurement equation;

[0026] Extended Kalman filter state equation:

[0027]

[0028] Where, X k is the state vector, f(X k )=[f1(X k ) f2(X k ) … f n (X k )] T is a nonlinear vector function, Γ k is the system noise allocation matrix, W k-1 is the system noise.

[0029] Extended Kalman filter measurement equation:

[0030]

[0031] In the formula, Z k is the system measurement vector, H k is the measurement matrix, V k is the measurement noise.

[0032] Further, the step S3 also includes: in the extended Kalman filter, two groups of covariance matrices of wheel odometers are provided respectively, and when the rubber-tyred trackless vehicle is moving, the first group of covariance data R1 is taken; when the rubber-tyred trackless vehicle is parked, the second group of covariance data R2 is taken (it should be noted that the covariance matrix of the wheel odometer should be reasonably determined according to the actual sensor type and actual working conditions, and only one example is provided here.):

[0033]

[0034] Furthermore, the step S4 also includes: establishing a strapdown inertial navigation / dead reckoning combined positioning model based on an improved adaptive extended Kalman filter algorithm to achieve secondary fusion of the encoder and the strapdown inertial navigation;

[0035] The improved adaptive extended Kalman filter algorithm is to improve the accuracy and stability of filtering by estimating the measurement noise covariance matrix through a given state noise covariance matrix;

[0036] The improved adaptive extended Kalman filter algorithm formula is as follows:

[0037] Filter calculation circuit:

[0038]

[0039] In the formula, is the one-step predicted value of the state, is the optimal estimate of the state, K k is the filter gain,

[0040] e kis the measurement prediction error, r k-1 =E[V k-1 ] is the mean value of the measurement noise.

[0041] Gain calculation loop:

[0042]

[0043] P k =(IK k H k ) k / k-1

[0044] Where P k / k-1 is the state one-step prediction mean square error matrix, P k is the mean square error matrix of the optimal state estimation, Q k is the measurement noise covariance matrix, R k is the measurement noise covariance matrix.

[0045] Adaptive Estimator:

[0046]

[0047] In the formula, is the correction factor, b is the fading factor, and its value range is (0,1).

[0048] Furthermore, the step S5 further comprises: the combined positioning fault-tolerant model adopts a residual χ 2 The inspection method detects and isolates the abnormal driving condition of the rubber-tyred trackless vehicle. The specific steps of this method are as follows:

[0049] In the extended Kalman filter process, the actual measurement value Z k and measured predicted values The difference between Defined as the residual, its expression is as follows:

[0050]

[0051] When the rubber-tyred trackless vehicle does not experience abnormal driving conditions, the residual It should satisfy the zero-mean Gaussian white noise distribution, and its variance σ k for:

[0052]

[0053] Where P k / k-1 is the one-step prediction mean square error matrix of the extended Kalman filter state, R k is the measurement noise covariance matrix;

[0054] When the rubber-tyred trackless vehicle encounters an abnormal driving condition, the residual It no longer conforms to the zero-mean Gaussian white noise distribution; through the residual The statistical characteristics of the trackless rubber-tyred vehicle are used to determine whether the vehicle has an abnormal driving condition.

[0055] The construction system detection function is:

[0056]

[0057] In the formula, λ k is subject to χ with m degrees of freedom 2 distribution, i.e., λ k ~x 2 (m), m is the measurement dimension;

[0058] like It is considered that the rubber-tyred trackless vehicle has an abnormal driving condition; if It is considered that the running condition of the rubber-tyred trackless vehicle is normal;

[0059] The detection threshold is set. When the system detects that the rubber-tyred trackless vehicle has an abnormal driving condition at a certain moment, the wheel odometer measurement information at that moment is considered abnormal, so the current measurement information is isolated and only the status is updated.

[0060] Furthermore, step S6 also includes: determining the transportation route of the rubber-tyred trackless vehicle by establishing a distribution map of underground coal mine tunnels, predetermining the location information of each shelter chamber in the transportation route, and arranging a QR code containing the location information of the shelter chamber at each shelter chamber.

[0061] Furthermore, step S7 also includes: when the rubber-tyred trackless vehicle enters each section and recognizes the QR code, the indicator light at the starting point of the section turns green, and the green arrow indicates the direction of travel. Before entering the next section, the green light remains on, and the indicator light at the end of the section remains in a gray state.

[0062] Furthermore, a multi-source information fusion positioning system for a trackless rubber-tyred vehicle in an underground coal mine is designed, comprising an encoder module, a strapdown inertial navigation module, a camera module, a bottom-level controller, an explosion-proof industrial computer, a ground PC and an explosion-proof power supply; the strapdown inertial navigation module is fixedly connected to the trackless rubber-tyred vehicle; the encoder modules are respectively installed on the non-steering wheels of the trackless rubber-tyred vehicle; the camera module is installed on the head of the trackless rubber-tyred vehicle; the bottom-level controller is respectively connected to the strapdown inertial navigation module and the encoder module, and is used for strapdown inertial navigation and encoder raw data and preprocessing; the explosion-proof industrial computer is respectively connected to the bottom-level controller and the camera module, and is used for multi-sensor data and multi-source information fusion positioning solution to execute the multi-source information fusion positioning method for trackless rubber-tyred vehicles in underground coal mines;

[0063] The ground PC is networked with the explosion-proof industrial computer and is used to receive and visualize the real-time positioning information of the rubber-tyred trackless vehicle obtained by the explosion-proof industrial computer; the explosion-proof power supply is connected to each module of the positioning system and supplies power.

[0064] Beneficial effects: The present invention provides a multi-source information fusion positioning system and method for a trackless rubber-tyred vehicle in an underground coal mine. The kinematic model of the wheel odometer of the trackless rubber-tyred vehicle is established by an encoder module installed on the non-steering wheel of the trackless rubber-tyred vehicle, and the forward speed and heading angle of the trackless rubber-tyred vehicle are output. The position, speed and posture of the trackless rubber-tyred vehicle are solved by a strapdown inertial navigation module fixed on the body of the trackless rubber-tyred vehicle. The heading angle solved by the wheel odometer kinematic model and the heading angle solved by the strapdown inertial navigation are fused by an extended Kalman filter to obtain the fused heading angle information. The odometer information of the wheel odometer and the fused heading angle information form a dead reckoning algorithm based on multiple encoders and strapdown inertial navigation, and the primary fusion of the encoder and the strapdown inertial navigation is realized. A strapdown inertial navigation / dead reckoning combined positioning model is established based on an improved adaptive extended Kalman filter algorithm, and the secondary fusion of the encoder and the strapdown inertial navigation is realized. A fault-tolerant model is established to realize the detection and isolation of abnormal driving conditions such as slipping or sliding of the trackless rubber-tyred vehicle.

[0065] In order to solve the problem of cumulative error in long-term and long-distance operation of the combined positioning model, several QR codes containing location information are posted in the shelter chamber in the underground coal mine tunnel, and the transportation route of the trackless rubber-tyred vehicle is divided into several driving sections using the location information set by the QR code. The QR code in the tunnel is scanned and identified by the camera module installed on the trackless rubber-tyred vehicle, so as to realize the staged calibration and reset of the positioning information of the trackless rubber-tyred vehicle. The multi-source information fusion positioning system and method have the advantages of high positioning accuracy, strong autonomy, and low cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 It is the hardware framework of the positioning system of the present invention;

[0067] Figure 2 It is a kinematic model of rubber-tyred trackless vehicle based on multiple encoders;

[0068] Figure 3 is a flow chart of the positioning method of the present invention;

[0069] Figure 4 This is the flow chart of the improved adaptive extended Kalman filter algorithm;

[0070] Figure 5 This is a schematic diagram of position calibration based on monocular camera recognition of QR code;

[0071] Figure 6This is a schematic diagram of the visualization of the real-time position information of the rubber-tyred trackless vehicle on the ground PC control platform. DETAILED DESCRIPTION

[0072] The present invention will be further explained below in conjunction with the accompanying drawings.

[0073] like Figure 1 As shown, it shows the logical relationship and connection relationship between the hardware modules of the positioning system. The hardware modules include encoder module, strapdown inertial navigation module, camera module, bottom controller, explosion-proof industrial computer, ground PC and explosion-proof power supply; the strapdown inertial navigation module is fixedly connected to the trackless rubber-wheeled vehicle; the encoder module is respectively installed on the non-steering wheel of the trackless rubber-wheeled vehicle; the camera module is installed on the head of the trackless rubber-wheeled vehicle; the bottom controller is respectively connected to the strapdown inertial navigation module and the encoder module, which is used for strapdown inertial navigation and encoder raw data and pre-processing; the explosion-proof industrial computer is respectively connected to the bottom controller and the camera module, which is used for multi-sensor data and multi-source information fusion positioning solution; the ground PC is networked with the explosion-proof industrial computer, which is used to receive the real-time positioning information of the trackless rubber-wheeled vehicle solved by the explosion-proof industrial computer and visualize it; the explosion-proof power supply is connected to each module of the positioning system and supplies power.

[0074] like Figure 2 As shown in the figure, the kinematic model of the wheel odometer established based on the encoder data of the non-steering wheel of the rubber-tyred trackless vehicle is:

[0075] First, by reading the accumulated number of pulses of the wheel encoder in a sampling period, the rotation speed of the left and right wheels of the trackless rubber-tyred vehicle can be calculated.

[0076]

[0077] Where N is the number of pulses accumulated in one rotation of the wheel encoder, M is the number of pulses measured in one sampling period, d is the diameter of the wheel, and T is the sampling period of the wheel encoder.

[0078] Then, the overall forward speed v of the rubber-tyred trackless vehicle can be calculated from the speeds of the left and right wheels of the trackless rubber-tyred vehicle. D , which is equal to the average of the speeds of the left and right wheels

[0079]

[0080] In the formula, v A is the speed of the left wheel of the rubber-tyred trackless vehicle, v B is the speed of the right wheel of the rubber-tyred trackless vehicle.

[0081] Secondly, according to the geometric relationship of the kinematic model of the trackless rubber-tyred vehicle, the heading angle change ψ of the trackless rubber-tyred vehicle can be calculated

[0082]

[0083] Where ψ is the change in the heading angle of the rubber-tyred trackless vehicle at two adjacent moments, L is the distance between the left and right wheels, and Δt is the time difference between two adjacent moments.

[0084] Finally, by differentiating the heading angle change ψ, we can get the heading angle change rate of the trackless rubber-tyred vehicle, that is, the angular velocity ω of the trackless rubber-tyred vehicle moving around the center point of the connecting axis of the two non-steering wheels

[0085]

[0086] Where ω is the angular velocity of the rubber-tyred trackless vehicle moving around the center point of the connecting axis of the two non-steering wheels.

[0087] Arranging the above equations, we can get the output of the kinematic model of the wheel odometer of the trackless rubber-wheeled vehicle based on multiple encoders:

[0088] [v D ψω] T

[0089] like Figure 3 FIG. 1 is a flow chart of the positioning method, which includes the following steps:

[0090] S1: A kinematic model of a wheel odometer of a rubber-tyred trackless vehicle is established based on the encoder data installed on the non-steering wheel of the rubber-tyred trackless vehicle. The kinematic model can be used to calculate the forward speed, heading angle and heading angular velocity information of the rubber-tyred trackless vehicle.

[0091] S2: Perform strapdown inertial navigation calculation according to the strapdown inertial navigation module data to obtain the attitude (including heading angle), speed and position information of the rubber-tyred trackless vehicle;

[0092] S3: The extended Kalman filter algorithm is used to fuse the heading data calculated based on the wheel odometer kinematic model and the heading angle calculated by the strapdown inertial navigation system. The dead reckoning algorithm based on the multi-encoder and strapdown inertial navigation system is obtained from the wheel mileage information and the fused heading angle information.

[0093] S4: A strapdown inertial navigation / dead reckoning combined positioning model is established based on an improved adaptive extended Kalman filter algorithm to achieve secondary fusion of encoder and strapdown inertial navigation;

[0094] S5: Establish a combined positioning fault-tolerant model to detect and isolate abnormal driving conditions such as slipping or sliding of rubber-tyred trackless vehicles, and improve the anti-interference ability of the combined positioning system;

[0095] S6: Several QR codes containing location information are posted in the shelter chambers in the underground tunnels of the coal mine, and the transportation route of the rubber-tyred trackless vehicle is divided into several driving sections using the location information set by the QR codes. The QR codes in the tunnels are scanned and identified using a camera module installed on the rubber-tyred trackless vehicle, so as to realize the staged calibration and resetting of the positioning information of the rubber-tyred trackless vehicle.

[0096] S7: Using the relative positioning information based on the combination of multiple encoders and strapdown inertial navigation depth, and the absolute positioning information obtained by scanning and recognizing the QR code with a monocular camera, the local relative positioning + global absolute positioning of the rubber-tyred trackless vehicle in the underground tunnel of the coal mine is realized.

[0097] Furthermore, the step S2 also includes: the strapdown inertial navigation solution algorithm is as follows:

[0098] Quaternion attitude update algorithm:

[0099]

[0100] In the formula, t k The quaternion of the carrier's attitude change at each moment, For n series from t k-1 to k The quaternion of the attitude change at the moment, t k-1 The quaternion of the attitude change of system b relative to system n at the moment, For b series from t k-1 to k Quaternion of attitude change at a given moment.

[0101] Speed ​​update algorithm:

[0102]

[0103] In the formula, and t k-1 and t k The inertial navigation solution speed at the moment, and They are update cycles T k =t k -t k-1 The velocity increment of the specific force velocity increment and the velocity increment of the harmful acceleration in the internal navigation coordinate system.

[0104] Location update algorithm:

[0105]

[0106] In the formula, p k and p k-1 tk and t k-1 The inertial navigation solution position at the time,

[0107]

[0108] Furthermore, the step S3 further comprises: k y k z k θ k γ k ψ k ] T (k represents t k The time (time) is defined as the state variable of the extended Kalman filter, the posture data calculated by the wheel odometer kinematic model is used as the prediction model to establish the extended Kalman filter state equation, and the posture data calculated by the strapdown inertial navigation system is used as the measurement model to establish the extended Kalman filter measurement equation;

[0109] Extended Kalman filter state equation:

[0110]

[0111] Where, X k is the state vector, f(X k )=[f1(X k )f2(X k )…f n (X k )] T is a nonlinear vector function, Γ k is the system noise allocation matrix, W k-1 is the system noise.

[0112] Extended Kalman filter measurement equation:

[0113]

[0114] In the formula, Z k is the system measurement vector, H k is the measurement matrix, V k is the measurement noise.

[0115] Further, the step S3 also includes: in the extended Kalman filter, two groups of covariance matrices of wheel odometers are provided respectively, and when the rubber-tyred trackless vehicle is moving, the first group of covariance data R1 is taken; when the rubber-tyred trackless vehicle is parked, the second group of covariance data R2 is taken (it should be noted that the covariance matrix of the wheel odometer should be reasonably determined according to the actual sensor type and actual working conditions, and only one example is provided here.):

[0116]

[0117] Further, the step S4 also includes: the improved adaptive extended Kalman filter algorithm refers to improving the accuracy and stability of filtering by estimating the measurement noise covariance matrix through a given state noise covariance matrix;

[0118] The improved adaptive extended Kalman filter algorithm formula is as follows:

[0119] Filter calculation circuit:

[0120]

[0121] In the formula, is the one-step predicted value of the state, is the optimal estimate of the state, K k is the filter gain,

[0122] e k is the measurement prediction error, r k-1 =E[V k-1 ] is the mean value of the measurement noise.

[0123] Gain calculation loop:

[0124]

[0125] Where P k / k-1 is the state one-step prediction mean square error matrix, P k is the mean square error matrix of the optimal state estimation, Q k is the measurement noise covariance matrix, R k is the measurement noise covariance matrix.

[0126] Adaptive Estimator:

[0127]

[0128] In the formula, is the correction factor, b is the fading factor, and its value range is (0,1).

[0129] Furthermore, the step S5 further comprises: the combined positioning fault-tolerant model adopts a residual χ 2 The inspection method detects and isolates the abnormal driving condition of the rubber-tyred trackless vehicle. The specific steps of this method are as follows:

[0130] In the extended Kalman filter process, the actual measurement value Z k and measured predicted values The difference between Defined as the residual, its expression is as follows:

[0131]

[0132] When the rubber-tyred trackless vehicle does not experience abnormal driving conditions, the residual It should satisfy the zero-mean Gaussian white noise distribution, and its variance σ k for:

[0133]

[0134] Where P k / k-1 is the one-step prediction mean square error matrix of the extended Kalman filter state, R k is the measurement noise covariance matrix;

[0135] When the rubber-tyred trackless vehicle encounters an abnormal driving condition, the residual It no longer conforms to the zero-mean Gaussian white noise distribution; through the residual The statistical characteristics of the trackless rubber-tyred vehicle are used to determine whether the vehicle has an abnormal driving condition.

[0136] The construction system detection function is:

[0137]

[0138] In the formula, λ k is subject to χ with m degrees of freedom 2 distribution, i.e., λ k ~x 2 (m), m is the measurement dimension;

[0139] like It is considered that the rubber-tyred trackless vehicle has an abnormal driving condition; if It is considered that the running condition of the rubber-tyred trackless vehicle is normal;

[0140] The detection threshold is set. When the system detects that the rubber-tyred trackless vehicle has an abnormal driving condition at a certain moment, the wheel odometer measurement information at that moment is considered abnormal, so the current measurement information is isolated and only the status is updated.

[0141] Furthermore, step S6 also includes: determining the transportation route of the rubber-tyred trackless vehicle by establishing a distribution map of underground coal mine tunnels, predetermining the location information of each shelter chamber in the transportation route, and arranging a QR code containing the location information of the shelter chamber at each shelter chamber.

[0142] Furthermore, step S7 also includes: when the rubber-tyred trackless vehicle enters each section and recognizes the QR code, the indicator light at the starting point of the section turns green, and the green arrow indicates the direction of travel. Before entering the next section, the green light remains on, and the indicator light at the end of the section remains in a gray state.

[0143] like Figure 4As shown in the figure, it is a flow chart of an improved adaptive extended Kalman filter algorithm, and its specific calculation formula is as follows:

[0144] Filter calculation circuit:

[0145]

[0146] In the formula, is the one-step predicted value of the state, is the optimal estimate of the state, K k is the filter gain,

[0147] e k is the measurement prediction error, r k-1 =E[V k-1 ] is the mean value of the measurement noise.

[0148] Gain calculation loop:

[0149]

[0150] Where P k / k-1 is the state one-step prediction mean square error matrix, P k is the mean square error matrix of the optimal state estimation, Q k is the measurement noise covariance matrix, R k is the measurement noise covariance matrix.

[0151] Adaptive Estimator:

[0152]

[0153] In the formula, is the correction factor, b is the fading factor, and its value range is (0,1).

[0154] like Figure 5 As shown in the figure, it is a schematic diagram of position calibration based on monocular camera recognition of QR code. The transportation route of the rubber-tyred trackless vehicle is determined by establishing a distribution map of underground coal mine tunnels, and the position information of each shelter chamber in the transportation route is determined in advance. A QR code containing the position information of the shelter chamber is arranged at each shelter chamber, and the transportation route of the rubber-tyred trackless vehicle is divided into several driving sections using the position information set by the QR code. The camera module installed on the rubber-tyred trackless vehicle is used to scan and identify the QR code in the tunnel to realize the staged calibration and reset of the positioning information of the rubber-tyred trackless vehicle.

[0155] like Figure 6As shown in the figure, it is a visualization diagram of the real-time position information of the rubber-tyred trackless vehicle on the ground PC control platform. When the rubber-tyred trackless vehicle enters each section and recognizes the QR code, the indicator light at the starting point of the section turns green, and the green arrow indicates the direction of travel. Before entering the next section, the green light remains on, and the indicator light at the end of the section (the starting point of the next section) remains in gray.

[0156] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A multi-source information fusion positioning method for a trackless rubber-tyred vehicle in a coal mine, characterized by: The method comprises the following steps: S1: A kinematic model of a wheel odometer of a rubber-tyred trackless vehicle is established based on the encoder data installed on the non-steering wheel of the rubber-tyred trackless vehicle, and the forward speed, heading angle and heading angular velocity information of the rubber-tyred trackless vehicle are obtained by solving the kinematic model; S2: Perform strapdown inertial navigation calculation based on the strapdown inertial navigation module data to obtain the attitude, speed and position information of the rubber-tyred trackless vehicle; S3: The extended Kalman filter algorithm is used to perform the first fusion of the heading data calculated based on the wheel odometer kinematic model and the heading angle calculated by the strapdown inertial navigation system. The dead reckoning algorithm based on the multi-encoder and strapdown inertial navigation system is obtained from the wheel odometer information and the fused heading angle information. S4: Establish a strapdown inertial navigation / dead reckoning combined positioning model to achieve the second fusion of encoder and strapdown inertial navigation; S5: Establish a combined positioning fault-tolerant model to detect and isolate abnormal driving conditions of rubber-tyred trackless vehicles and improve the anti-interference ability of the combined positioning system; S6: a plurality of QR codes containing position information are set at the vehicle shelter chamber in the underground coal mine tunnel, and the transportation route of the trackless rubber-tyred vehicle is divided into a plurality of driving sections using the position information set by the QR code, and the QR code in the tunnel is scanned and identified using a camera module installed on the trackless rubber-tyred vehicle to achieve staged calibration and resetting of the positioning information of the trackless rubber-tyred vehicle; S7: Using the relative positioning information based on the combination of multi-encoders and strapdown inertial navigation depth, and the absolute positioning information obtained by scanning and recognizing the QR code with a monocular camera, the local relative positioning and global absolute positioning of the rubber-tyred trackless vehicle in the underground lanes of the coal mine are combined; The step S3 further comprises: k y k z k θ k γ k ψ k ] T Defined as the state variable of the extended Kalman filter, the posture data calculated by the wheel odometer kinematic model is used as the prediction model to establish the extended Kalman filter state equation, and the posture data calculated by the strapdown inertial navigation system is used as the measurement model to establish the extended Kalman filter measurement equation; Extended Kalman filter state equation: Where, X k is the state vector, f(X k )=[f1(X k ) f2(X k )…f n (X k )] T is a nonlinear vector function, Γ k-1 is the system noise allocation matrix, W k-1 is the system noise; Extended Kalman filter measurement equation: In the formula, Z k is the system measurement vector, H k is the measurement matrix, V k is the measurement noise; The step S3 further includes: in the extended Kalman filter, two groups of covariance matrices of wheel odometers are provided respectively, and when the rubber-tyred trackless vehicle is moving, the first group of covariance data R1 is taken; when the rubber-tyred trackless vehicle is parked, the second group of covariance data R2 is taken: The step S4 also includes: establishing a strapdown inertial navigation / dead reckoning combined positioning model based on an improved adaptive extended Kalman filter algorithm to achieve secondary fusion of the encoder and the strapdown inertial navigation; The improved adaptive extended Kalman filter algorithm is to improve the accuracy and stability of filtering by estimating the measurement noise covariance matrix through a given state noise covariance matrix; The improved adaptive extended Kalman filter algorithm formula is as follows: Filter calculation circuit: In the formula, is the one-step predicted value of the state, is the optimal estimate of the state, K k is the filter gain, e k is the measurement of prediction error, Gain calculation loop: P k =(I-K k H k )P k / k-1 Where P k / k-1 is the state one-step prediction mean square error matrix, P k is the mean square error matrix of the optimal state estimation, Q k is the measurement noise covariance matrix, R k is the measurement noise covariance matrix; Adaptive Estimator: In the formula, is the correction factor, b is the fading factor, and its value range is (0,1) The step S5 also includes: the combined positioning fault-tolerant model adopts a residual χ 2 The inspection method detects and isolates the abnormal driving condition of the rubber-tyred trackless vehicle. The specific steps of this method are as follows: In the extended Kalman filter process, the actual measurement value Z k and measured predicted values The difference between Defined as the residual, its expression is as follows: When the rubber-tyred trackless vehicle does not experience abnormal driving conditions, the residual It should satisfy the zero-mean Gaussian white noise distribution, and its variance σ k for: Where P k / k-1 is the one-step prediction mean square error matrix of the extended Kalman filter state, R k is the measurement noise covariance matrix; When the rubber-tyred trackless vehicle encounters an abnormal driving condition, the residual It no longer conforms to the zero-mean Gaussian white noise distribution; through the residual The statistical characteristics of the trackless rubber-tyred vehicle are used to determine whether the vehicle has an abnormal driving condition. The construction system detection function is: In the formula, λ k is subject to χ with m degrees of freedom 2 distribution, i.e., λ k ~x 2 (m), m is the measurement dimension; like It is considered that the rubber-tyred trackless vehicle has an abnormal driving condition; if It is considered that the running condition of the rubber-tyred trackless vehicle is normal; The detection threshold is set. When the system detects that the rubber-tyred trackless vehicle has an abnormal driving condition at a certain moment, the wheel odometer measurement information at that moment is considered abnormal, so the current measurement information is isolated and only the status is updated.

2. A method for multi-source information fusion positioning of a trackless rubber-tyred vehicle in a coal mine according to claim 1, characterized in that: The step S6 also includes: determining the transportation route of the rubber-tyred trackless vehicle by establishing a distribution map of underground coal mine tunnels, predetermining the location information of each shelter chamber in the transportation route, and arranging a two-dimensional code containing the location information of the shelter chamber at each shelter chamber.

3. A method for multi-source information fusion positioning of a trackless rubber-tyred vehicle in a coal mine according to claim 2, characterized in that: The step S7 also includes: when the rubber-tyred trackless vehicle enters each section and recognizes the QR code, the indicator light at the starting point of the section turns green, and the green arrow indicates the direction of travel. Before entering the next section, the green light remains on, and the indicator light at the end of the section remains in a gray state.

4. A multi-source information fusion positioning system for trackless rubber-tyred vehicles in coal mines, characterized in that: The invention comprises an encoder module, a strapdown inertial navigation module, a camera module, a bottom controller, an explosion-proof industrial computer, a ground PC and an explosion-proof power supply; the strapdown inertial navigation module is fixedly connected to the trackless rubber-wheeled vehicle; the encoder modules are respectively installed on the non-steering wheels of the trackless rubber-wheeled vehicle; the camera module is installed on the head of the trackless rubber-wheeled vehicle; the bottom controller is respectively connected to the strapdown inertial navigation module and the encoder module, and is used for strapdown inertial navigation and encoder raw data and preprocessing; the explosion-proof industrial computer is respectively connected to the bottom controller and the camera module, and is used for multi-sensor data and multi-source information fusion positioning and solving to execute the multi-source information fusion positioning method for trackless rubber-wheeled vehicles in coal mines as claimed in claim 1; the ground PC is networked with the explosion-proof industrial computer, and is used for receiving the real-time positioning information of the trackless rubber-wheeled vehicle solved by the explosion-proof industrial computer and visualizing it; the explosion-proof power supply is connected to each module of the positioning system and supplies power.

Images