Summary of the invention
The present invention is directed to complete a set of intelligent vehicle navigation system based on perception and from host computer location navigation, which can
To apply in the occasion for needing to realize that article or personnel transport automatically, such as factory, office building, family's environment.By GPS or
After the modes such as RFID give initial position and the direction of vehicle, the number of a variety of multiple sensors such as movement, ranging, image is utilized
According to, be compared and calculate with information such as architecture structure drawing or electronic maps, thus complete dynamic calibration and determine car body on ground
Position on figure.Cooperate the functional modules such as motor driven, wireless communication, image recognition, motion-sensing simultaneously, completes each wheel
Positioning respectively and control, and then realize intelligent vehicle positioning, path planning, move freely, intelligent barrier avoiding, be in communication with each other
Function.This system can be by network real-time monitoring and the distribution by wireless communication of the intelligent terminals such as mobile phone, tablet computer, desktop computer
Task realizes the visualization and real-time control of intelligent vehicle system.
In order to solve the above-mentioned technical problem, a kind of intelligence based on perception and from host computer location navigation proposed by the present invention
Vehicle navigation system, including the information storage module, motor drive module, locating module, sense being connected respectively with central processing unit
Know module and interactively communicate module, the central processing unit, drive module and locating module for realizing intelligent vehicle positioning
Navigation and movement, the sensing module is for realizing the steady control of intelligent vehicle, and the module that interactively communicates is for realizing people
Machine interaction;The steering of the intelligent vehicle uses differential steering, and the movement of each wheel of intelligent vehicle is respectively by independent electricity
Machine control, the motor is stepper motor or direct current generator, and the driving of each motor is all made of motor drive module control;It is described
Locating module includes at least by motion sensor means, ranging sensing device, image sensing device, RFID device and GPS system;Often
The motion sensor means and ranging sensing device are separately installed at a wheel, the data acquisition of each wheel is relatively independent;
The RFID device includes RFID reception module and multiple RFID transmitting modules, and the RFID reception module is arranged on vehicle,
Multiple RFID transmitting modules are arranged in the different location in building;The sensing module for realizing intelligence speed perception,
Face perception and height change perception, in which: the speed perception is realized using coding disk or speedometer motion-sensing element, is used for
The velocity of rotation of wheel is perceived, to realize the control and speed feedback control of intact stability;The ground perception is using figure
As sensor or infrared sensor realization, material and flatness for perceiving ground;The height change perception is using acceleration
Degree meter motion sensor is realized, so that the evenness of road surface situation of each wheel region of intelligent vehicle is judged, to be wheel
Control realize data support;The module that interactively communicates includes communication submodule and human-computer interaction submodule, wherein communicator
Module is responsible for transformation task data and operating state data, is also used for the communication between remote control terminal and central processing unit
In swapping status data between multiple intelligent vehicle systems, to prevent route from clashing;The human-computer interaction is from module
Including I/O unit, control and the setting of parameter are carried out to intelligent vehicle system for artificial scene;The central processing unit
For collecting other all module data collected, at the same to motor drive module, interactively communicate module and send corresponding control
Information processed, and pass through human-computer interaction submodule to user's output information.
The process of navigation is realized using the above-mentioned intelligent vehicle navigation system based on perception and from host computer location navigation, including
Following steps:
Step 1: the acquisition of broad sense electronic map: broad sense electronic map refers to that is electronically stored is used for intelligent vehicle
System indoors occasion realize autonomous positioning and navigation architecture structure drawing and intelligent vehicle outdoor occasion realize autonomous positioning and
The electronic map of navigation;Barrier, marker, movement limitation, pavement behavior and calibration point are also recorded on the electronic map;
Before work of navigating, broad sense electronic map is imported in the information storage module of intelligent vehicle system, and electronically according to broad sense
Figure and the corresponding relationship of physical location establish absolute coordinate system, read and utilize when to run;
Step 2: the acquisition of vehicle initial position message and positioning are calibrated, vehicle initial position message uses GPS or RFID
Method obtains, all the sensors in the locating module position from host computer, calculates vehicle within a certain period of time
Displacement and the direction of motion, and separated in time or apart from using GPS or RFID approach and by outside positioning datum come
The current location for correcting vehicle, to carry out dynamic calibration;At the same time, using on architecture structure drawing complaint message and channel
Information limits the possible direction of motion of intelligent vehicle and range in a certain range, and the history road passed by according to intelligent vehicle
Diameter information carries out continuous processing to the real time position of vehicle, obtains the possibility model of intelligent vehicle position within a unit time
It encloses;
Step 3: obtaining the current position coordinates and coordinates of targets of vehicle: the current position coordinates of vehicle pass through initial bit
Confidence is ceased and is obtained from the result of host computer positioning and dynamic calibration, and coordinates of targets is given by top level control system according to actual needs
It is fixed;
Step 4: Dynamic Programming travel route: after obtaining initial position and target position, to architecture structure drawing or electricity
After sub- map carries out corresponding standardization, route planning is carried out using two-way A* algorithm, genetic algorithm and neural network, when
When vehicle location and preset travel route deviate more than 2-10m, planned again according to current position coordinates and target location coordinate
Route realizes the Dynamic Programming of route;Policy selection distance minimal path, most steady route and the fastest road of route planning
One of line;Dynamic Programming travel route realizes the overall traveling control of vehicle, wherein cooperation wheel perception realizes that vehicle exists
Travel route controls in regional area.
Further, the present invention is based on perception and from the intelligent vehicle navigation procedure of host computer location navigation, in step 4
Involved wheel perception includes perception, the perception on ground and the perception of height change of wheel velocity, in which:
The perception of wheel velocity is to adjust driving signal in time according to the difference of actual rotation speed and pre-set velocity, makes reality
Border velocity of rotation lifts a ban preset value as far as possible, to realize the closed-loop control of speed;
The perception on ground is the image for acquiring vehicle front in real time by imaging sensor, gives central processing unit according to figure
As Processing Algorithm is handled, whether identification vehicle operation front ground is smooth, what material of ground shifts to an earlier date to control the system
Take corresponding control measure;For the ground of out-of-flatness, vehicle is determined according to the area in out-of-flatness region and degree of roughness
Whether detour;The coefficient of friction on wheel and ground is estimated according to the identification of material, to take different strategy of speed control;
The perception of height change is the perception of wheel pitch amplitude, when vehicle driving to roughness pavement is, setting
The output of accelerometer at wheel will appear violent fluctuation, and system takes deceleration to handle accordingly, to reduce pitch
Influence to vehicle judges road over there according to the distribution for the wheel that jolts if only pitch occurs in piece wheel
Road is more smooth, and then guides vehicle to smooth direction running.
Compared with prior art, the beneficial effects of the present invention are:
Compared to existing AGV system, the present invention has following points advantage:
First, the present invention uses the localization method calculated based on architecture structure drawing or electronic map.This positioning method is not
Electromagnetic wave is relied on, not will receive electromagnetic interference.Only by the calculating of the sensing devices such as the movement of itself completion displacement etc., and according to
The information that the dynamic calibrations data comparison such as RFID builds mechanism map or electronic map etc. completes the calculating of position, accomplishes based on
The independent positioning of calculation.Since architecture structure drawing and electronic map etc. are all existing information, it is also required to be surveyed in advance
The work such as draw.Have the characteristics that it is portable strong, using simple and quick.
Second, the present invention has intelligent Path selection and planning function.Most intelligent vehicles are all using pre- on the market at present
It is first laid with route pattern on the ground, route and tracking traveling are then identified in the way of image procossing etc..If met pre-
If there is the case where obstacle etc. cannot move on route, can only stop working.And the present invention need to only be pre-loaded into interior architecture
Structure chart or electronic map, and it is equipped with avoidance and obstruction warning system, the autonomous route planning and navigation of vehicle can be realized, no
It is limited by pre-determined route, is worked together more intelligent and efficient.
Third realizes the independent perception and control of wheel.Limit of traditional intelligent vehicle system by route, routing algorithm etc.
System, tends to completion task but road pavement situation is without adaptibility to response, fortune is easily caused there are non-smooth-riding surface
Send the damage of article.The present invention in several ways perceives wheel and travel condition of vehicle, realizes regional area
Route optimization and control optimization, thus more steady, safety transport personnel and cargo.
Specific embodiment
Technical solution of the present invention is described in further detail in the following with reference to the drawings and specific embodiments, it is described specific
Embodiment is only explained the present invention, is not intended to limit the invention.
A kind of intelligent vehicle navigation system based on perception and from host computer location navigation proposed by the present invention, including distinguish
The information storage module that is connected with central processing unit, motor drive module, locating module, sensing module and module is interactively communicated,
It is of course also possible to which other function module is added as needed.
The central processing unit, motor drive module and locating module are basic function modules, for realizing intelligent vehicle
Location navigation and movement.
Respectively by independent motor control, the motor is stepper motor or direct current for the movement of each wheel of intelligent vehicle
Motor, the driving of each motor are all made of motor drive module control.Comprehensively consider the stationarity and steering characteristic of vehicle, car body
It is designed using four-wheel, size is determined by actual needs.The power resources of vehicle use the design of stepper motor or direct current generator, and
Do so it is environmentally protective, do not cause exhaust emissions etc. pollute.The movement of each one wheel of motor control.The design of more motors can
To guarantee that even running and the movement including turning to of vehicle accurately control.Motor driven uses motor drive module, such as
The characteristics of L298N has driving capability strong, and calorific value is low, strong antijamming capability.For using direct current generator, L298N electricity
Machine drive module can drive two direct current generators simultaneously, i.e. L298N motor drive module can be inputted from IN1, IN2 respectively
Pwm signal drives the revolving speed and direction of a motor, similarly can drive another electricity from IN3, IN4 input pwm signal respectively
The revolving speed of machine and direction.The turn-around design of the intelligent vehicle uses the scheme of differential steering.So-called differential refers to left and right two
The speed difference of wheel, if left side wheel is faster than the right, then trolley can be biased to the right side.Conversely, left wheel rotating ratio right wheel turns
Fast slow, that one trolley can be rotated to the left side.Therefore, the motor on control left side right side is rotated and reverse respectively with identical speed,
It can realize that vehicle turns at any angle.
The locating module includes at least by motion sensor means, ranging sensing device, image sensing device, RFID device
And GPS system;Special emphasis be sensing device used in locating module include but is not limited to motion sensor means (such as
Speedometer, accelerometer, gyroscope and various novel sports sensors etc.), ranging sensing device (such as infrared range-measurement system, swash
Optar, ultrasonic distance measuring apparatus and various novel distance measuring sensors etc.), image sensing device (such as camera etc.) etc..Each
The motion sensor means and ranging sensing device are separately installed at wheel, the data acquisition of each wheel is relatively independent.Institute
Stating RFID device includes RFID reception module and multiple RFID transmitting modules, carries out vehicle initial position using the method for RFID
It determines and is only applicable to indoor environment with positioning calibration, therefore, the RFID reception module is arranged on vehicle, multiple RFID transmittings
Module is also installed only in building, in order to realize dynamic calibration function, the different location in building is needed to install RFID
Transmitting module.The specific installation site of RFID transmitting module generally by positioning accuracy requirement and actual building structure it is true
It is fixed.Higher to positioning accuracy request from interval, then the spacing between RFID transmitting module should be smaller, in list
Reach higher calibration number in the time of position.In addition, should also be pacified at the entrance in channel, room etc. according to building structure
RFID transmitting module is filled, because these positions can provide a range for the position of personnel.
The sensing module is for realizing the steady control of intelligent vehicle, and the sensing module is for realizing intelligence speed sense
Know, ground perception and height change perception etc. wherein: the speed perception is the velocity of rotation for perceiving wheel, so as to reality
The control and speed feedback control of existing intact stability, the function can be real with the motion-sensings such as coding disk or speedometer element
It is existing.The material etc. of flatness of the ground perception for perceiving ground, can use imaging sensor or infrared sensor is real
It is existing.The height change perception refers to that the car body as caused by ground or other factors pitches, by being distributed in each wheel
Height change induction element obtain be which (or which) wheel region pavement roughness, thus be wheel control
System realizes that data are supported, the perception of height change can be realized by motion sensors such as accelerometers.Except imaging sensor belongs to
It is shared with other systems, other sensors should all be mounted at each wheel, and the acquisition of data is relatively independent.
The module that interactively communicates is for realizing human-computer interaction.The module that interactively communicates includes communication submodule and man-machine
Interaction submodule, wherein communication submodule is responsible for transformation task number for the communication between remote control terminal and central processing unit
According to and operating state data, be also used to swap status data between multiple intelligent vehicle systems, to prevent route from rushing
It is prominent;The human-computer interaction includes I/O unit (keyboard, Mike, display screen, loudspeaker etc.) from module, is used for manually existing
Field carries out control and the setting of parameter to intelligent vehicle system.
The central processing unit for collecting other all module data collected, while to motor drive module, hand over
Mutual communication module sends corresponding control information, and passes through human-computer interaction submodule to user's output information according to positioning result.
Realize that the process of navigation is main using the above-mentioned intelligent vehicle navigation system based on perception and from host computer location navigation
It is that location navigation algorithm is written in central processing unit completing above-mentioned hardware system with postponing.System is before use, need to make
With the corresponding architecture structure drawing in place or electronic map information import system, so that the positioning and route planning of system operate.System
After system starting, since the design is used based on the location algorithm calculated with autonomous reckoning positioning, so first having to by artificial or phase
The positioning datum answered gives the initial position message where intelligent vehicle, as the benchmark positioned later.Then, it is calculated by route planning
Method cooks up the route of vehicle projected trip using architecture structure drawing or electronic map information and destination coordinate.It finally transfers to drive
Dynamic system executes the mobile route, and real-time calculating location information and feeds back in moving process, until the task of completion.Specific step
It is rapid as follows:
Step 1: the acquisition of broad sense electronic map: broad sense electronic map refers to that is electronically stored is used for intelligent vehicle
System indoors occasion realize autonomous positioning and navigation architecture structure drawing and intelligent vehicle outdoor occasion realize autonomous positioning and
The electronic map etc. of navigation;In computer widely applied today, this type of information is easily obtained.In addition, can also exist as needed
The information such as barrier, marker, movement limitation, pavement behavior and calibration point are also recorded on the electronic map;In navigation work
Before work, broad sense electronic map information should be imported in the information storage module of intelligent vehicle system, and according to broad sense electronic map
Absolute coordinate system is established with the corresponding relationship of physical location, reads and utilizes when to run.
Step 2: the acquisition of vehicle initial position message and positioning are calibrated, it is based on calculating since this system takes one kind
Autonomous positioning mode, so initial position message must be obtained before navigation by starting autonomous positioning in system.The comprehensive present invention
Use scope and use environment, this system can take following two mode to obtain initial position message.
One is the methods based on satellite positioning.This method makes full use of existing satellite positioning network, such as GPS
The big-dipper satellite of (Global Satellite Positioning, GPS), China's independent research positions system
System etc..This method has the advantage that locating speed is fast, overlay area is wide, but covering poor indoor environment to satellite-signal can not
Reach good effect, is mainly used in outdoor environment in this way.It needs to install additional in intelligent vehicle system using this method
Corresponding satellite positioning signal receiving module.
Another kind is the method using RFID (Radio Frequency Identification, radio frequency identification).It uses
This method needs to lay a certain number of RFID transmitting modules according to actual needs in the range of needing and positioning in advance, and in electricity
Its position is marked on sub- cartographic information, and corresponding receiving module is installed in intelligent vehicle system.It can be believed in this way by radio frequency
Number the information such as intensity judge current location.
It is of the invention in the specific implementation, should at least be used according to practical service environment above one or more of modes into
Line home position information gives.
The information of the various kinds of sensors in the locating module carried due to the present invention using system itself carries out autonomous
Positioning is calculated, the displacement and the direction of motion of vehicle within a certain period of time are calculated, so position error or the accumulation with the time
And it is stepped up.So separated in time or distance is needed just to calibrate to positioning.The method of calibration i.e. use GPS or
RFID method and be to correct the current location of vehicle by external positioning datum, to carry out dynamic calibration;Its method can
To obtain identical method using with initial position message.
Various kinds of sensors of the data source that real time position calculates in locating module, including but not limited to motion sensor
(such as speedometer, accelerometer, gyroscope and various novel sports sensors etc.), range unit (such as infrared range-measurement system, swash
Optar, ultrasonic distance measuring apparatus and various novel sports sensors etc.), image sensing device (camera etc.) etc..By above-mentioned
The data of sensor can calculate the displacement of object within a certain period of time by certain method.
Such as according to the definition of acceleration it is recognised that speed is integral of the accelerometer to the time, and being displaced is speed pair
The integral of time.I.e. acceleration, which carries out integral twice to the time, can calculate the displacement of object.So passing through accelerometer
Real time acceleration value is obtained, then it is done by processor and is integrated twice, the distance of displacement can be calculated.Meanwhile according to
The directions such as gyroscope measure device, the angular speed of real-time detection object of which movement, and calculate the direction of motion of object in real time.
Ranging sensing device can realize calculating and the basis of position by identification object to the distance of fixed reference
Calculated result carries out dynamic calibration.So-called fixed reference has the object of fixed position, the position of these objects on cartographic information
Fixation is set, will not be changed easily, such as the objects such as door, window, wall in building.
At the same time, using the complaint message and channel information and other necessary informations on architecture structure drawing, it will be intelligent
The possible direction of motion of vehicle and range limit in a certain range, and the historical path information passed by according to intelligent vehicle is to vehicle
Real time position carry out continuous processing, obtain the possible range of intelligent vehicle position within a unit time;
In this way, the displacement of intelligent vehicle relative initial position can be learnt using motion sensor, it can using distance measuring sensor
Measure intelligent vehicle arrive fixed reference (such as wall) distance, in conjunction with by architecture structure drawing and historical path acquisition letter
Breath more can accurately determine the position of intelligent vehicle.
Step 3: current position coordinates and coordinates of targets are the necessary informations for route planning and navigation.Current location
Coordinate information can be positioned by initial position message, from host computer and the information acquisitions such as dynamic calibration.Coordinates of targets then may be used
To be given according to actual needs by top level control system.
Step 4: Dynamic Programming travel route: after obtaining initial position and target position, route planning can work
Give location algorithm.After carrying out corresponding standardization to architecture structure drawing or electronic map, that is, it can be used at present
More mature two-way A* algorithm, genetic algorithm, neural network etc. carry out route planning, deviate (partially when vehicle location is biggish
When from being greater than 2-10m) preset travel route when, according to current position coordinates and target location coordinate again programme path, realize
The Dynamic Programming of route;The policy selection distance minimal path of route planning, most steady route (avoid known bumpy road and barrier
One of hinder) and fastest route etc..
In the present invention, Dynamic Programming travel route realizes the overall traveling control of vehicle, and wheel perception is cooperated to realize vehicle
Travel route controls in regional area.In a variety of application modes (such as large cargo transport, personnel's transport) of the invention,
Car body size is larger, therefore, is not able to satisfy the needs to motion control to the location navigation precision of intelligent vehicle system entirety.Cause
This, the positioning to intelligent vehicle system entirety not only may be implemented in the present invention, can also realize in moving process to each wheel
Perception respectively and control.Therefore, the present invention is each wheel in design using the control of independent motor drive module, so
Possibility is provided for the independent control of wheel.It should be pointed out that being a kind of partial zones by the control that wheel perception is realized
The control in domain, the route totally travelled should also follow the data of path planning navigation module.
Perceiving data obtained can bundle with position data, and architecture structure drawing or electronics are added to after processing
In the information of map, thus the process of path optimizing planning.
The purpose of the independent control of wheel is in the state of guaranteeing steady intelligent vehicle system, safety, as far as possible quickly
Completion task.
It is the control to speed first, speed has important influence to the steady and safety that vehicle is run.It is each first
The precision of a wheel velocity control.The driving signal issued by drive module is possible to because of extraneous factor the control of motor
It influences and generates difference, cause the speed difference between each wheel, unstable, direction deviation, vehicle are run so as to cause vehicle
The consequences such as foot wheel abrasion.The perception of speed is exactly the closed-loop control for speed, thus according to actual rotation speed and pre-set velocity
Difference adjust driving signal in time, so that actual rotation speed is lifted a ban preset value as far as possible, to realize the closed-loop control of speed.
The perception on ground is the image for acquiring vehicle front in real time by imaging sensor, gives central processing unit according to figure
As Processing Algorithm is handled, whether identification vehicle operation front ground is smooth, is what material etc., shifts to an earlier date to control the system
Take corresponding control measure;For the ground of out-of-flatness, vehicle is determined according to the area in out-of-flatness region and degree of roughness
Whether detour;The coefficient of friction on wheel and ground is estimated according to the identification of material, to take different strategy of speed control;
All roughness pavements are not identified out using the method for image procossing, so just needing wheel while there is height
The function of degree variation (wheel pitch amplitude) perception.When vehicle driving to roughness pavement is the biography being arranged at wheel
The output of sensor (such as accelerometer) will appear violent fluctuation, and accordingly, central controller can take corresponding measure.Such as,
Deceleration processing, to reduce influence of the pitch to transport cargo or personnel.For another example, if only piece wheel jolts,
It can judge that road over there is more smooth according to the distribution for the wheel that jolts, and then guide vehicle to smooth direction running.
The perception of above-mentioned wheel is only intended to route and the control optimization of regional area with control, whole route planning with lead
Boat is still realized with path planning algorithm.
Although above in conjunction with attached drawing, invention has been described, and the invention is not limited to above-mentioned specific implementations
Mode, the above mentioned embodiment is only schematical, rather than restrictive, and those skilled in the art are at this
Under the enlightenment of invention, without deviating from the spirit of the invention, many variations can also be made, these belong to of the invention
Within protection.