Summary of the invention
The purpose of the present invention is to propose to a kind of, and the coal mine fully-mechanized mining face based on SLAM technology is maked an inspection tour unmanned vehicle and is independently patrolled
Depending on method, the high error of positioning accuracy is small.
The technical scheme is that unmanned vehicle, the fully mechanized coal face are maked an inspection tour in a kind of coal mine fully-mechanized mining face based on SLAM technology
Making an inspection tour unmanned vehicle includes binocular camera, protective frame, chassis, explosion-proof tank, it is characterised in that: also equipped with intelligence on the unmanned vehicle
It can module and sensor module;The intelligent object includes coal mine environment image processing module, binocular camera demarcating module, comprehensive
Adopt face SLAM module, the maintenance of coal mine map and unmanned vehicle path planning module;The explosion-proof tank is welded on unmanned chassis, institute
It states binocular camera to be mounted on roof, protective frame is installed on binocular camera top;The sensor module includes inertia sensing
Device;
The motion information of the inertial sensor acquisition unmanned vehicle;
The binocular camera obtains the image information of unmanned vehicle;
The coal mine environment image processing module is by the practical mark in binocular camera acquired image information and mine
Determine environment and establishes mapping relations;
The fully mechanized coal face SLAM module is collected according to binocular camera captured image information and inertial sensor
Motion information positions position of the unmanned vehicle in mine, creates initial coal mine environmental map;
The unmanned vehicle that mine map maintenance and unmanned vehicle path planning module obtain fully mechanized coal face SLAM module is in mine
After position location information and initial coal mine Environmental Map Information in well are ceaselessly updated, system is controlled by unmanned vehicle
System and the optimal path that other positions in unmanned vehicle arrival mine are determined using the path planning algorithm of rolling window.
Unmanned vehicle, which is maked an inspection tour, the present invention is based on the coal mine fully-mechanized mining face of SLAM technology realizes that the method independently maked an inspection tour includes following step
It is rapid:
Step 1: the intrinsic parameter of calibration binocular camera, initializes fully mechanized coal face SLAM module, and obtain initial
Coal mine environment map;
Step 2: unmanned vehicle mine make an inspection tour when, binocular camera will current location acquired image information and just
Beginning coal mine environment map carries out Feature Points Matching and carries out third step if successful match;If matching is unsuccessful, unmanned vehicle
Control system just will be considered that current fully mechanized coal face make an inspection tour unmanned vehicle position alignment quality it is poor, then can trigger fully mechanized coal face and make an inspection tour nobody
The image information that binocular camera obtains, i.e., is added to the initial coal mine environmental map of the first step by the part reorientation work of vehicle
In, then Feature Points Matching is re-started, until successful match;
Step 3: if the success of local map Feature Points Matching, unmanned vehicle are established using the ICP algorithm based on least square method
Fully mechanized coal face make an inspection tour unmanned vehicle motion model, using SVD decomposition method solve motion model, in conjunction with nonlinear optimization mode solve it is comprehensive
Adopt the pose of face unmanned vehicle;Unmanned vehicle constantly adds the environmental map point of new position during tour simultaneously, constitutes current coal
Mine environmental map, while current coal mine environment map is updated into ceaselessly row;
Step 4: passing through unmanned vehicle control after the completion of unmanned vehicle is to the corresponding tour task in current mine position
And determine that unmanned vehicle reaches the optimal path of next target position in mine using the path planning algorithm of rolling window, nobody
Vehicle will be reached next target position with optimal path and carry out continuing to make an inspection tour work.
It is multiple in coal mine fully-mechanized mining face, coal mine hydraulic supporting, tunnel etc. to unmanned vehicle that the present invention is based on vision SLAM technologies
Position carries out autonomous positioning and environment construction, while unmanned vehicle can carry out patrol monitoring, the condition of a disaster early warning to entire mine, can be with
The actual change situation for making staff easily grasp fully-mechanized mining working very much and geographical location and ring below entire mine
Border variation, and the work such as can give warning in advance to the condition of a disaster, largely save manpower and financial resources.
Specific embodiment
It illustrates, specifically includes described below in the following, unmanned vehicle will be maked an inspection tour to the application fully mechanized coal face with reference to attached drawing:
It is as shown in Figure 1 that fully mechanized coal face makes an inspection tour the specific contour model schematic diagram of unmanned vehicle.Including binocular camera 2, protective frame 1,
Chassis 4, explosion-proof tank 3, the explosion-proof tank 3 are welded on fully mechanized coal face and make an inspection tour on unmanned chassis 4, and the binocular camera 2 is mounted on
On roof, protective frame 1 is installed on 2 top of binocular camera, prevents coal cinder or other objects from tumbling and smashing binocular camera.
The unmanned vehicle customization explosion-proof tank 3 uses aluminum alloy casing die cast, surface high pressure electrostatic plastic spraying.
As shown in Fig. 2, the fully mechanized coal face is maked an inspection tour on unmanned vehicle also equipped with intelligent object and inertial sensor;The intelligence
Module include coal mine environment image processing module, binocular camera demarcating module, fully mechanized coal face SLAM module, mine map maintenance with
Unmanned vehicle path planning module.
The unmanned vehicle control is made of tall and handsome up to TX2 exploitation version and ROS operating system.Trolley is grasped using ROS
The communication between different hardware is realized as system.The development kit that ROS system is provided using official develops version up to TX2 tall and handsome
Upper creation ROS topic, by fully mechanized coal face make an inspection tour the control algolithm of unmanned vehicle, fully mechanized coal face SLAM algorithm, rolling window path planning
Algorithm etc. is written in development board in advance.
As shown in figure 3, making an inspection tour unmanned vehicle the present invention is based on the fully mechanized coal face of SLAM technology realizes autonomous positioning, navigation, mine
The technical problems such as fully mechanized coal face tour are through the following steps that complete.
The first step, due to fully mechanized coal face make an inspection tour unmanned vehicle working environment particularity, first to unmanned vehicle carry binocular take the photograph
As head does intrinsic parameter calibration, fully mechanized coal face SLAM module is initialized, and obtain initial coal mine environmental map.
Specific step is as follows for intrinsic parameter calibration:
Binocular camera is directed at the fully mechanized coal face environment templates to be demarcated by step 1);
Step 2) starts binocular camera demarcating module, is guaranteeing that the coal mine fully-mechanized mining face environment templates to be demarcated of unmanned vehicle are complete
Under the premise of portion is located in the binocular camera field of view of unmanned vehicle carrying, from least two directions to the fully mechanized coal face to be demarcated
Environment templates are taken pictures, and calibrating template image is acquired;
Step 3) is corresponding with the practical calibration characteristic point foundation of environment by the characteristic point in collected demarcating module image
Relationship calculates the intrinsic parameter square that fully mechanized coal face makes an inspection tour the camera optimization of unmanned vehicle using the plane reference algorithm of Zhang Zhengyou automatically
Battle array, and it is saved in binocular camera demarcating module.
Second step, unmanned vehicle, will be in the collected coal mine environment image informations in current location and initial when mine is maked an inspection tour
Coal mine environment map carries out Feature Points Matching and carries out third step if successful match;If matching is unsuccessful, unmanned vehicle
Control system just will be considered that the position alignment quality of current fully mechanized coal face tour unmanned vehicle is poor, then can trigger fully mechanized coal face and make an inspection tour unmanned vehicle
Part reorientation work.
Part reorientation work specifically includes the following steps:
Step 1): the initial coal mine environment that SLAM module is established is added in the collected topography's information of binocular camera
In map;
Step 2): using binocular camera that acquired image information is corresponding with unmanned vehicle current working site new
Environmental map carries out Feature Points Matching.
Step 3: if Feature Points Matching success, unmanned vehicle are established fully mechanized coal face using the ICP algorithm based on least square method and are patrolled
Depending on unmanned vehicle motion model, solve motion model using SVD decomposition method, in conjunction with nonlinear optimization mode solve fully mechanized coal face nobody
The current pose of vehicle;Fully mechanized coal face SLAM module is automatically updated according to matching result simultaneously;It updates fully mechanized coal face and makes an inspection tour unmanned vehicle current
The map feature point of operating position, and the updated map feature point is added in initial coal mine environmental map, update is worked as
Preceding unmanned vehicle is in the corresponding map in coal mine position.
The present invention uses grey level histogram least mean-square error (the Minimum Mean based on coal mine environment principal direction
Squared Error, initialism MMSE) algorithm progress Feature Points Matching.
4th step passes through unmanned vehicle control after the completion of unmanned vehicle is to the corresponding tour task in current mine position
And determine that unmanned vehicle reaches the optimal path of next target position in mine using the path planning algorithm of rolling window, nobody
Vehicle will be reached next target position with optimal path and carry out continuing to make an inspection tour work.
Specific step is as follows for the path planning algorithm of rolling window:
Step 1) initializes all parameters of unmanned vehicle, such as starting point, target point, working environment, and rolling window is arranged
Size, fully mechanized coal face make an inspection tour the step-length that unmanned vehicle moves every time;
Step 2) is continuously updated the environmental information of current scrolling window, if there are dynamic barrier, roots in window
According to the motion state of the subsequent time of its linear prediction model prediction barrier;If without dynamic barrier, mine unmanned vehicle
According to the local optimum specific item punctuate generated in window, the planning one local optimum path mobile towards its direction;
Step 3) makes an inspection tour unmanned vehicle to fully mechanized coal face and makes corresponding Robot dodge strategy according to prediction of collision;
Step 4) fully mechanized coal face makes an inspection tour unmanned vehicle according to corresponding Robot dodge strategy, stops a period of time or change speed to find out one
Item and the collisionless path of barrier;
Step 5) is according to the step number of setting, and to the path movement planned, and step-length cannot be greater than windows radius;
If step 6) fully mechanized coal face makes an inspection tour unmanned vehicle and reaches target position, algorithm terminates, and the above content is combine specific side
Formula is described in detail the present invention.
Although preferred embodiments of the present invention have been described, it is created once a person skilled in the art knows basic
Property concept, then additional changes and modifications may be made to these embodiments.So it includes excellent that the following claims are intended to be interpreted as
It selects embodiment and falls into all change and modification of the scope of the invention.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.In this way, if these modifications and changes of the present invention belongs to the range of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to include these modifications and variations.