CN102745196A - Intelligent control device and method for granular computing-based micro intelligent vehicle - Google Patents
Intelligent control device and method for granular computing-based micro intelligent vehicle Download PDFInfo
- Publication number
- CN102745196A CN102745196A CN2012102494806A CN201210249480A CN102745196A CN 102745196 A CN102745196 A CN 102745196A CN 2012102494806 A CN2012102494806 A CN 2012102494806A CN 201210249480 A CN201210249480 A CN 201210249480A CN 102745196 A CN102745196 A CN 102745196A
- Authority
- CN
- China
- Prior art keywords
- control
- parameter
- rule
- information
- corner
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Traffic Control Systems (AREA)
Abstract
The embodiment of the invention provides an intelligent control device for a micro intelligent vehicle. The intelligent control device comprises a camera (10), a signal detection module (20), a control rule module (31), a control parameter calculation module (32), an Arduino control panel (33), a motor driver (40) and a steering engine driver (50). The invention also provides an intelligent control method for the micro intelligent vehicle. The control rule is acquired by employing the granular computing theory, the micro intelligent vehicle is intelligently controlled, and a control method for traditionally controlling a precise mathematical model required to be established, controlling in real time by establishing hierarchical control rules and gradually fining is avoided. The intelligent control device has the advantages of high control precision, high real-time property and high interference resistance.
Description
Technical field
The present invention relates to intelligent transportation field, use the grain theory of computation and automatic control theory, a kind of miniature intelligent vehicle intelligence controlling device and method of calculating based on grain is provided.
Background technology
Along with economic with social developing rapidly; The 'bottleneck' restrictions effect of traffic infrastructure is more and more obvious; This restriction not only is embodied in traffic blocking problem and becomes increasingly conspicuous, and also is embodied in problem of environmental pollution that does not freely cause owing to traffic and the road that falls behind relatively and the vehicle formed potential safety hazard of life, property to people simultaneously, Just because of this; Intelligent transportation system (Intelligent Transportation Systems is called for short ITS) receives the concern of associated mechanisms day by day.
ITS mainly comprises two parts: intelligent vehicle and intelligent traffic control system.Intelligent vehicle is a system ensemble that integrates environment sensing, programmed decision-making, multistage driver assistance function, and its core missions are to realize the automatic guidance of intelligent vehicle.And realize that the related gordian technique of intelligent vehicle automatically controlled intelligent traffic control system mainly comprises environment sensing, navigator fix, path planning, Decision Control technology etc.Wherein the Decision Control technology is equivalent to the brain of intelligent vehicle, and its radical function is according to the information that sensory perceptual system the obtains judgement of making a strategic decision, and carries out next step behavior then and makes a strategic decision, thereby reach the effect to the real-time control of intelligent vehicle.
For the Decision Control of miniature intelligent vehicle, prior art adopts tradition control and Based Intelligent Control usually.So-called tradition control; What it was representative is PID control (PID Control, i.e. ratio, integration, differential control), and traditional control need be set up a precise math model; And miniature intelligent vehicle is the system of a complicacy; Have very strong time variation and non-linear, be difficult to set up precise math model with traditional control method, and the poor anti jamming capability of tradition control.So-called Based Intelligent Control is that the experience with the people is that prerequisite is summarized a cover control law with knowledge, realizes its Based Intelligent Control according to this, and Based Intelligent Control need not set up precise math model, and it is stronger to control antijamming capability than traditional PID controller.Like: publication number is that patent of invention " speed tracking fuzzy control method of vehicle of vehicle drive robot ", traffic and transportation system engineering and the information of CN 101667015A is in April, 2010; The 10th the 2nd phase of volume; " based on the autonomous running method research of the intelligent vehicle of fuzzy control " of publication etc. have adopted fuzzy control method to realize independently going of intelligent car.Fuzzy control method is to be the intelligence control method on basis with the fuzzy theory, and its antijamming capability is strong, and need not set up precise math model.But fuzzy control method is under the traffic environment of complicacy, and the control law of summarizing can produce the situation of " blast ", thereby can reduce the automatically controlled real-time performance.
Summary of the invention
To miniature intelligent vehicle in the face of complicated traffic environment the time; Traditional PID controlling method is difficult to set up precise math model and poor anti jamming capability; And fuzzy control method is when facing complicated traffic environment, and the rule of formulation can produce " blast " and cause the relatively poor problem of control real-time, the present invention proposes a kind of intelligence controlling device and method based on grain calculating; The present invention need not set up precise math model, has stronger antijamming capability, in real time controller performance is better.
Miniature intelligent vehicle intelligence controlling device of the present invention comprises camera 10, signal detection module 20, control law module 31, controlled variable computing module 32, Arduino control desk 33, motor driver 40 and steering wheel actuator 50, it is characterized in that:
Said signal detection module 20 detects road condition information from the information that camera 10 obtains; Said road condition information is any one or the multiple combination arbitrarily in obstacle information, traffic lights information, taper mark information, lane mark information, the traffic sign information;
Said control law module 31 comprises the control law acquiring unit 31A and real-time control unit 31B that are connected through memory device 31C;
Said control law acquiring unit 31A; Be used to obtain control law; Memory device 31C is used for the control law that storage rule acquiring unit 31A obtains; Said control rule are according to acquisition testing road condition information and current speed parameter and the corner parameter of miniature intelligent vehicle, and the application grain theory of computation obtains after making up decision table;
Said control law acquiring unit 31A at first, obtains the rule of control rate; Secondly, be not under 0 the situation in speed, obtain the rule of control corner.Wherein, the rule of obtaining the control corner comprises the rule of obtaining the control direction again, obtain the rule of control between corner regions, obtain rule 3 parts of the concrete angle of control.
Said real-time control unit 31B; Detection road condition information that is used for signal detection module 20 is obtained and the control law of memory device 31C mate, and export the result and the cooresponding coverage of all control laws that are complementary and give controlled variable computing module 32.
Said controlled variable computing module 32, result and cooresponding coverage computation speed parameter and corner parameter according to all control laws that mate send to Arduino control desk 33 with speed parameter and corner parameter;
Said Arduino control desk 33 converts speed parameter to motor control signal, and motor control signal is sent to motor driver 40, and the corner Parameters Transformation is become the steering wheel control signal, and the steering wheel control signal is sent to steering wheel actuator 50.
Miniature intelligent vehicle intelligence control method of the present invention comprises:
11, utilize camera collection road conditions signal;
12, from the road conditions signal, detect road condition information, comprise any one or multiple combination arbitrarily in obstacle information, traffic lights information, taper mark information, lane mark information, the traffic sign information;
13, mate road condition information and control law, export the control law information of all couplings, comprise the result and the cooresponding coverage of the control law of coupling;
14, according to the control law information calculations speed parameter and the corner parameter of all couplings, speed parameter and corner parameter are sent to the Arduino control desk;
15, the Arduino control desk converts speed parameter to speed governor control signal respectively, and the corner Parameters Transformation is become the steering wheel control signal.
Than prior art; The present invention uses the Based Intelligent Control that a theory of computation realizes miniature intelligent vehicle; The precise math model of having avoided traditional control to set up; Through setting up the control law of level, progressively add smart control method in control in real time, have the better controlling precision, real-time is good and antijamming capability is strong.
Description of drawings
Fig. 1 is miniature intelligent vehicle modular construction figure;
Fig. 2 is miniature intelligent vehicle model;
Fig. 3 the present invention is based on the miniature intelligent vehicle control setup preferred embodiment constructional drawing that grain calculates;
Fig. 4 is that the miniature intelligent vehicle that the present invention is based on grain calculating obtains control law preferred embodiment diagram of circuit;
Fig. 5 the present invention is based on the miniature intelligent vehicle intelligence control method preferred embodiment diagram of circuit that grain calculates;
The specific embodiment
In order to make the object of the invention, technical scheme and advantage clearer, the present invention is explained further details below in conjunction with accompanying drawing.
It is the simulating human thinking and the new mode of dealing with problems in the current computational intelligence field of research that grain calculates, and is the effective tool of complex problem solving.Present main grain computation model has speech computation model, rough set model, quotient space model and other model etc.People in the face of complicated, be difficult to the problem accurately held, be not usually the employing system, accurate method goes the optimum solution of the problem of pursuing, but reaches limited reasonable target through the way of progressively attempting, and just obtains so-called enough separating of satisfaction.The mankind adopt multi-granularity analysis method this summary, from coarse to fine, continuous refinement, have avoided the high difficulty of computation complexity.
The present invention combines grain to calculate and Intelligent Control Theory, has proposed a kind of miniature intelligent vehicle intelligence control method that calculates based on grain.The present invention has stronger antijamming capability than traditional PID controller control, has stronger real-time control ability than fuzzy control simultaneously.
Miniature intelligent vehicle structural framing according to the invention intelligent vehicle model as shown in Figure 1, miniature is as shown in Figure 2, generally includes camera 10, X86 mainboard 1, Arduino control desk 33; And motor driver 40 and steering wheel actuator 50, its camera comprises camera and following camera two parts, said X86 mainboard 1 is this area general mainboard; Obviously; It will be apparent to those skilled in the art that miniature intelligent vehicle according to the invention also can adopt other mainboards to realize, no longer details.
The miniature intelligent vehicle intelligence controlling device of the present invention; Preferred embodiment is as shown in Figure 3, comprises camera 10, signal detection module 20, control law module 31, controlled variable computing module 32, Arduino control desk 33, motor driver 40 and steering wheel actuator 50.
Said signal detection module 20 comprises any one or the multiple combination arbitrarily in detection of obstacles unit 21, traffic lights detecting unit 22, taper mark detecting unit 23, lane mark detecting unit 24 and the traffic sign detecting unit 25; Said detection of obstacles unit 21, traffic lights detecting unit 22, taper mark detecting unit 23, lane mark detecting unit 24 and traffic sign detecting unit 25 detect obstacle information, traffic lights information, taper mark information, lane mark information and traffic sign information respectively from the information that camera 10 transmits.
Said detection of obstacles unit 21 mainly detects the distance of obstacle orientation, road the place ahead (comprising dead ahead, left front, right front), obstacle and miniature intelligent vehicle.
Said traffic lights detecting unit 22, whether have traffic lights, be which kind of lamp if mainly detecting the place ahead.
Said taper mark detecting unit 23, whether mainly detect road the place ahead has the taper mark, and the taper mark is put angle, and the distance of taper mark and miniature intelligent vehicle.
Said lane mark detecting unit 24 mainly detects the angle between road grade and the fare.
Said traffic sign detecting unit 25 mainly detects traffic prohibitory sign and speed of a motor vehicle sign.
Said signal detection module 20 partial functions can adopt image processing software to carry out; Representative type; Like OpenCV; It is the computer vision storehouse of increasing income that is provided by Intel Company, is made up of a series of C functions and a small amount of C++ class, has realized a lot of general-purpose algorithms of image processing and computer vision aspect.Typically, comprising:
Said detection of obstacles unit 21 utilizes OpenCV to extract the image of two camera collections up and down, and through gray scale and binary conversion treatment, the position deviation through image pixel has determined whether obstacle again.
Said traffic lights detecting unit 22 utilizes OpenCV to extract the image of camera collection, obtains triple channel RGB image, and R, the intensive place of G value accounting example have similar part with detection of obstacles in the detected image.
Said taper mark detecting unit 23 utilizes OpenCV to extract the image of camera collection, obtains triple channel RGB image, detects the continuous shared orientation of R colour, and it is fitted to straight line, calculates the angle that the taper mark is put again.
Said lane mark detecting unit 24; Utilize OpenCV to extract the image of camera collection, the single channel gray level image of acquisition road carries out the adaptive threshold binary conversion treatment to image; The single channel binary image is carried out rim detection; In through the image after the edge detection process, carry out hunting and handle, confirm the left-lane line or/and the right lane line, calculate the angle of lane mark and miniature intelligent vehicle;
Said traffic sign detecting unit 25, traffic sign is divided into stated-speed sign and road signal.General traffic sign is the sign that comprises some specific meanings with circle.Stated-speed sign identification and Traffic Sign Recognition need be first location all, find out the position and the size of the sign that possibly exist in the image.The sign of extraction circle the inside is being judged this sign then, thereby draws the traffic sign of concrete implication.
Miniature intelligent vehicle is when facing complicated traffic environment; Traditional PID controlling method is difficult to set up precise math model and poor anti jamming capability; And fuzzy control method is when facing complicated traffic environment, and the rule of formulation can produce " blast " and cause that to control real-time relatively poor.Thereby proposed a kind of Based Intelligent Control measure of calculating based on grain, it need not set up precise math model, has stronger antijamming capability, controller performance is better in real time.
Said control law module 31 can adopt existing control law to realize, such as directly obtaining speed parameter and corner parameter etc. according to traffic lights, record is to some extent repeated no more in existing document.
Preferably, control law module 31 according to the invention comprises the control law acquiring unit 31A and real-time control unit 31B that is connected through memory device 31C.
Said control law acquiring unit 31A according to the speed and the corner that detect road condition information and miniature intelligent vehicle, obtains control law and is stored among the memory device 31C, and is as shown in Figure 4, mainly comprises:
101, data acquisition
Miniature intelligent vehicle in motion, acquired signal detection module 20 detected obstacle informations, traffic lights information, taper mark information, lane mark information and traffic sign information, and miniature intelligent vehicle present speed parameter and corner parameter.
102, make up decision table, each is gathered constantly as delegation (delegation is defined as an object), with detection module 20 detected information and speed parameter and corner parameter as row, with the form preservation of form.
In the present embodiment, be conditional attribute with signal detection module 20 detected information definitions, speed parameter and corner parameter-definition are decision attribute, representative type, and as an instance, it is as shown in table 1 to make up decision table.
Table 1
Data capture definition is a collection moment each time, and sampling frequency can be provided with according to equipment with the collection duration flexibly, usually; Sampling frequency 5-10Hz gathered duration 0.5-2 hour, certainly; Also can adopt other sampling frequencys and gather duration, not have special qualification.
Preferably; As optional step, be included in after the data of gathering miniature intelligent vehicle preserve, data are carried out completeness detect; Promptly judge whether to collect wrong data; If wrong data produce, these misdatas can produce inconsistent data in decision table, then delete identical and the less object (OK) that decision attribute is different of conditional attribute in the decision table.Said inconsistent data is to have identical condition but decision attribute data inequality.
103, the definition of miniature intelligent vehicle grain and calculating
Because people's cognitive ability is limited; Old friends are in the face of complicated problems the time; At first complicated problems is divided into several better simply according to its characteristic and performance separately; Each that so marks off can be regarded an information soon as, that is to say, information is that feature similarity, function are close, the set of entity with fuzzy property.And just call the information granulation to the process that information is divided into information.
Generally speaking, information is divided into the two large divisions: the connotation and extension, intension are used for portrayal or represent a feature or function exactly; Extension then is illustrated in the set that has identical or close entity under this feature or function.
Briefly, grain calculating is exactly expression and process information grain.
Grain calculates the problem that mainly contains two aspects: the calculating of the definition of grain and use grain.
The grain definition (g representes with symbol) of miniature intelligent vehicle is g=((v, v among the present invention
D1, v
D2), m (v, v
D1, v
D2)), it divides the connotation and extension.Intension (v, v
D1, v
D2) constitute by three parts, v expression signal detection module 20 detected a kind of information are the value under the conditional attribute in the decision table, v
D1And v
D2Expression speed parameter and corner parameter, extension m (v, v
D1, v
D2) expression satisfies intension (v, v
D1, v
D2) all object sets.Wherein, (v, v
D1, v
D2) ∈ { * } ∪ V, * representes desirable arbitrary value.
The friendship computing formula of miniature intelligent vehicle grain is defined as:
G=g∧g’=((v,v
d1,v
d2),m(v，v
d1,v
d2))∧((v’,v
d1’,v
d2’),m(v’,v
d1’,v
d2’))
=((v∧v’,v
d1∧v
d1’,v
d2∧v
d2’),m(v,v
d1,v
d2)∧m(v’,v
d1’,v
d2’))
With table 1 is example, a given grain g=((b=0 ∧ c=1, d1=0, d2=0), T5, and T6}), the intension of this g is: traffic lights detecting unit 22 detected information are 0, and detection of obstacles unit 21 detected information are 1, and this moment, speed parameter was 0, and corner is 0; Extension then is: traffic lights detecting unit 22 detected information are 0, and detection of obstacles unit 21 detected information are 1, and this moment, speed parameter was 0, and corner is all set constantly, i.e. { T5, T6} of 0.
104, according to decision table, use the grain theory of computation, obtain control law, further comprise:
104-1 at first obtains the speed rule
The value of the value of each row and speed attribute in 104-1A, the ergodic condition attribute is placed on moment of identical value in the identity set, forms the set of different values under this conditional attribute.With the intension of the value under each row as grain, the cooresponding set of formation is as extension.
For example: the ergodic condition attribute is value and the set of formation thereof under the traffic lights (b), and cooresponding grain is like following table 2:
Table 2
Travel through the value under value under all conditions attribute and the speed attribute, obtain different value set under the different condition attribute, and cooresponding grain.Like following table 3:
Table 3
Cooresponding each set under 104-1B, the ergodic condition attribute, whether judgement wherein each speed parameter constantly is identical, if then obtain (the representing with cov) of the regular and corresponding set of value under this conditional attribute, and delete and should gather; Be about under the conditional attribute particle respectively with the speed attribute under cooresponding particle calculate according to the computing formula of miniature intelligent vehicle grain; Form new particle; Whether the confidence level of judging new particle again is 1; If be 1 then the intension of the new particle of formation is the control law of acquisition and (the representing with cov) of extension, and delete this particle.Wherein, the confidence calculations formula of particle be Conf (g)=| m (v, v
D1, *) |/| m (v, *, *) |.
With table 1, table 2 is example; Conditional attribute is that the value of traffic lights (b) is 0 o'clock set { T1, T3, T5, T6}; Inquiry T1, T3, T5, T6, T7 be cooresponding speed parameter constantly; Find that T1, T3, T5, T6, T7 speed parameter constantly are incomplete same, then can not obtain the rule of value under this conditional attribute;
Be that conditional attribute is that the value of traffic lights (b) is 0 o'clock cooresponding particle g1={ (0
b, *, *), (T1, T3, T5, T6, T7) } respectively with the speed attribute under particle g9={ (195
D1, *, *), (T1, T3, T4, T7) }, g10={ (0
D1, *, *), (T2, T5, T6) } ship and obtain:
G
1=g1∧g9={(0
b,195
d1,*)，(T1、T3、T7)}
G
2=g1∧g10={(0
b,0
d1,*)，(T5、T6)}
According to the confidence calculations formula, can get Conf (G
1)=| (T1, T3, T7) |/| (T1, T3, T5, T6, T7) |=0.6
Conf(G
2)=|(T5、T6)|/|(T1、T3、T5、T6、T7)|=0.4
Because Conf (G
1), Conf (G
2) confidence level be not 1, so can not under this conditional attribute value, obtain rule.
Conditional attribute is that the value of traffic lights (b) is that { T2}, speed parameter are identical (have only a value, be regarded as identical), then obtain the rule under the value 1 of traffic lights, and promptly during the value 1 of traffic lights, its velocity amplitude is 0, and deletion should set for 1 o'clock set;
Conditional attribute is that the value of traffic lights (b) is 1 o'clock cooresponding particle g2={ (1
b, *, *), (T2) } respectively with the speed attribute under particle g9={ (195
D1, *, *), (T1, T3, T4, T7) }, g10={ (0
D1, *, *), (T2, T5, T6) } ship and obtain:
G
3=g2∧g9={(1
b,195
d1,*)，()}
G
4=g2∧g10={(1
b,0
d1,*)，(T2)}
At this moment, Conf (G
1)=0, Conf (G
2)=1; G
2Intension can be used as a control law, promptly the traffic lights value is 1 o'clock, its speed is 0, deletion g2 and G
4
This rule simple table is shown as: if b=1then d1=0; Cov={T2} (for ease of describing, the present invention all representes rule with the procedural language mode with further part).
In like manner, obtain rule under the value 2 of traffic lights.
According to above process, under obstacle, fare angle attribute, obtain control law respectively.Finally the rule of the control rate of acquisition is as follows:
If?b=1then?d1=0;cov={T2}
If?b=2then?d1=195;cov={T4}
If?c=1then?d1=0;cov={T5，T6}
If?a=-17then?d1=195;cov={T7}
Residue set and cooresponding grain thereof are (seeing shown in the table 4):
Table 4
Judge whether all moment union of sets collection of obtaining rule equal all complete or collected works constantly, and { T2, T4, T5, T6, T7} are not equal to complete or collected works in the set in the moment of all rules that obtain at this moment.Continue execution in step 104-1C.
104-1C, calculation is shipped in the set of value under any two conditional attributes, obtain cooresponding common factor, i.e. calculating to any two particles utilization grain under the different condition attribute forms new particle, returns step 104B;
Promptly (1), be that { T1, T3, T5, T6, T7} are that { T1, T2, T3, T4, T7}, fare angle (a) value are that { T1, T6}, fare angle (a) value are that { T2, T4}, fare angle (a) value are that { T3, T5} seek common ground in 62 cooresponding set in-15 cooresponding set in 0 cooresponding set in 0 cooresponding set with conditional attribute obstacle (c) value respectively in 0 cooresponding set with traffic lights (b) value; (2), conditional attribute obstacle (c) value be 0 cooresponding set T1, T2, T3, T4, T7} respectively with fare angle (a) value be 0 cooresponding set T1, T6}, fare angle (a) value be-15 cooresponding set T2, T4}, fare angle (a) value be 62 cooresponding set T3, T5} seek common ground, and be g1 respectively with g4, g6, g7, g8; G4 obtains new particle with g6, g7, g8 according to the computing formula of grain respectively.See shown in the table 5:
Table 5
According to above 104-1B step; Conditional attribute traffic lights value be 0 and conditional attribute obstacle value be that { T1 among T1, the T3}, T3 speed constantly are identical for 0 set; Then obtain rule and be " when conditional attribute traffic lights value 0 and conditional attribute obstacle value 0, its speed is 195 ", and set { T1, the T3} of deletion conditional attribute traffic lights value 0 and conditional attribute obstacle value 0; If speed is inequality, then preserving should set.
That is, g1 and g8 utilization grain operation rule obtains G
5={ (0
b∧ 0
c, 195
D1, *), (T1, T3) }, confidence level is Conf (G
5)=1.
In like manner, obtain rule " the obstacle value be 0 and the fare angle be 0 o'clock, its speed is 195 " and " the obstacle value is that 0 o'clock and fare angle are 62 o'clock, and its speed is 195 ".
At this moment, obtain the rule as follows:
Ifb=0and?c=0then?d1=195cov={T1、T3、T7};
Ifb=0and?a=0then?d1=195cov={T1};
Ifb=0and?a=62then?d1=195cov={T3};
104-1D, when all moment union of sets collection of obtaining rule equal all constantly during complete or collected works, finish the also redundant rule of deletion, otherwise, return step 104-1C.
Whole rules that above-mentioned steps is obtained are following:
Rule (1): If b=1then d1=0; Cov={T2}
Rule (2): If b=2then d1=195; Cov={T4}
Rule (3): If c=1then d1=0; Cov={T5, T6}
Rule (4): If a=-17then d1=195; Cov={T7}
Rule (5): If b=0and c=0then d1=195; Cov={T1, T3, T7}
Rule (6): If b=0and a=0then d1=195; Cov={T1}
Rule (7): If b=0and a=62then d1=195; Cov={T3}
At this moment, the moment union of sets collection of strictly all rules equals all complete or collected works constantly, and the rule of obtaining control rate finishes.
If the method for deletion redundancy rule is judged the control law that is obtained, successively the cov of one of them control law
iBe another one control law cov
jSubclass should rule be redundancy rule then, need the deletion.
Obviously; Rule (4), rule (6), rule (7) cooresponding cov are the subclass of rule (5) cooresponding cov in the above-mentioned rule; Therefore, rule (4), rule (6), rule (7) are redundancy rules, deletion rule (4), rule (6), rule (7).
The rule of the final control rate that obtains is:
Rule (1): If b=1then d=0; Cov={T2}
Rule (2): If b=2then d=195; Cov={T4}
Rule (3): If c=1then d=0; Cov={T5, T6}
Rule (4): If b=0and c=0then d1=195; Cov={T1, T3, T7}
104-2 uses the grain theory of computation, obtains the rule of control corner;
The rule of obtaining the control corner under 0 the situation in speed.The rule of wherein obtaining the control corner comprises again:
(1) obtains the rule of controlling direction (turn left, keep straight on, turn right);
(2) under the situation of turning left or turning right, obtain the rule (this angular interval can freely be set at the 0-90 degree, and the present invention spends as an interval with 5) of controlling angular interval;
(3) obtain the control law of controlling concrete angle under the interval of each in (2).
More than table obtains control law for example, comprising:
104-2A obtains the rule of control direction (turn left, keep straight on, turn right)
When miniature intelligent vehicle in motion, if speed is not 0, then at first need judge direction: turn left, keep straight on, turn right.
The embodiment of the invention is because the angle of when operation, keeping straight on that miniature intelligent vehicle is set is 80; The angle of turning left>80, the angle of right-hand rotation 80, therefore; Be not the description decision-making that can increase a direction under 0 the situation in speed, the decision-making value is: 1 (left-hand rotation) ,-1 (right-hand rotation), 0 (craspedodrome).With table 1 is example, and table 1 medium velocity is not that 0 data can be converted into table 6:
Table 6
As shown in table 6, it is similar with 104-1A to obtain regular step, and the value of each row and direction are described the value of attribute in the ergodic condition attribute, and moment of identical value is placed in the identity set, forms the set of different values under this conditional attribute.With the intension of the value under each row as grain, the cooresponding set of formation is as extension.
Likewise, be similar to step 104-1B ~ 104-1D, the control law that obtains is following:
Rule (1): if a=0then direction=0; Cov={T1}
Rule (2): if a=62then direction=-1; Cov={T3}
Rule (3): if a=-15then direction=1; Cov={T4}
Rule (4): if a=-17then direction=1; Cov={T7}
Rule (5): if b=2then direction=1; Cov={T4}
Preferably; These control laws are merged; Judge these control laws under the identical situation of decision attribute value, can the value under the Rule of judgment attribute represent (or connect with ' or ') with an interval value, and cov then obtains with the union of sets computing; Thereby obtain the rule of lesser amt, controlled in real time in order to improve.Rule with above-mentioned control direction is an example, and rule (3), rule (4) and rule (5) are all to be 1 on the value of direction at decision attribute, and rule (3), rule (4) are inequality with rule (5) conditional attribute, so can not merge; Rule (3) can merge with rule (4), and the rule of merging is if-17 < a < 15then direction=1; Cov={T4, T7} (or if a=-17or a=-15then direction=1; Cov={T4, T7}).
Rule (1): if a=0then direction=0; Cov={T1}
Rule (2): if a=62then direction=-1; Cov={T3}
Rule (3): if a=-17or a=-15then direction=1cov={T4, T7}
Rule (4): if b=2then direction=1; Cov={T4}
Obviously, rule (4) is a redundancy rule, needs deletion.Need to prove,, the situation of rule (4) then can not occur being similar to when the data of gathering are complete more.
104-2B obtains the rule between the control corner regions
When miniature intelligent vehicle left or to the right the time, need further obtain angle to the left or to the right, be to improve antijamming capability, the present invention turn left or the right-hand rotation situation under obtaining the rule between the control corner regions earlier; And miniature intelligent vehicle has only an angle when keeping straight on, therefore not to obtaining the rule of controlling between corner regions under the craspedodrome situation.
The value of angular interval is a unit (can between 0-80, freely be provided with between this corner regions certainly) with 5 among the present invention; 80 < d2 ≤85 is a unit (being labeled as 1) under the left-hand rotation situation; 85 < d2 ≤90 is a unit (being labeled as 2) ..., 80+5* (i-1) < d2 ≤80+5*i; Be designated as i, handle successively; Equally, 75 ≤d2 < 80 is a unit (being labeled as 1) under the right-hand rotation situation, and 70 ≤d2 < 70 is a unit (being labeled as 2) ..., < < 80-5* (i-1) is designated as i to=d2 to 80-5*i, handles successively.
In above-mentioned table 6, the decision table under left-hand rotation or right-hand rotation situation is convertible into like following table 7, table 8:
Table 7
Table 8
To decision table 7, the decision table 8 after transforming, with the similar acquisition control law of step 104-1A ~ 104-1D.The value of the value of each row and interval mark attribute in the ergodic condition attribute is placed on moment of identical value in the identity set, forms the set of different values under this conditional attribute.With the intension of the value under each row as grain, the cooresponding set of formation is as extension.
104-2C obtains the rule of the concrete angle of control
On the basis of 104-2C, in all angles interval, obtain the rule of the concrete angle of control respectively, for example angular interval unit is 1 situation making policy decision table such as following table 9 in table 7:
Table 9
To decision table 9, with the similar acquisition control law of step 104-1A ~ 104-1D.The value of the value of each row and direction attribute in the ergodic condition attribute is placed on moment of identical value in the identity set, forms the set of different values under this conditional attribute.With the intension of the value under each row as grain, the cooresponding set of formation is as extension.
With the resulting control law of above step, deposit in as among the memory module 31C among Fig. 3.Storage mode can be stored with tabulated form and (or store with form of decision tree; When controlling in real time, can reduce the time of matched rule with this storage mode; More can improve the controlled in real time of miniature intelligent vehicle); The embodiment of the invention adopts the tabulated form storage, like the control rate rule that obtains in the above-mentioned example is:
（1）If?b=1then?d=0；cov={T2}
（2）If?b=2then?d=195；cov={T4}
（3）If?c=1then?d=0；cov={T5，T6}
（4）If?b=0and?c=0then?d1=195cov={T1，T3、T7}
File layout such as the table 10 of these rules in the control law module:
Table 10
Wherein * representes all possible value.
In like manner, can control the rule of direction, the rule between the control corner regions, the rule of the concrete angle of control with the stored in form of form respectively.
During control in real time; Gather traffic information by camera 10; Detection module 20 detects according to the traffic information of camera collection, and control module 31B matees according to the control law among detection module 20 detected traffic informations and the memory device 31C in real time, and exports the control law information that is complementary; Controlled variable computing module 32 is according to the control law information calculations speed parameter and the corner parameter of control module 31B output in real time, and concrete steps are following:
With control law in the memory device (31C) as condition; If the detected road condition information of signal detection module (20) satisfies the rule condition of control rate in the memory device (31C); Then export this (or many) control law information (cooresponding speed parameter of this rule or corner parameter and this regular coverage), these information spinners will be divided into three kinds of situation:
(1) rule that is not complementary, speed parameter are empty, and coverage is empty;
(2) have only a rule coupling, a speed parameter, coverage is the cooresponding coverage of matching rules;
(3) many rule couplings are arranged, one or more speed parameters, coverage is the cooresponding coverage of each bar control law.
Controlled variable computing module 32 according to the control law information calculations speed parameter of real-time control unit 31B output, and sends to Arduino control desk 33 with speed parameter.
According to three kinds of situation control law information of output, accordingly, controlled variable computing module 32 has three kinds to handle processing method:
Situation 1: speed remains unchanged, and sends the speed parameter of last time and gives Arduino control desk 33.Especially, when the speed of last time was 0, same or transmission speed parameter was given Arduino control desk 33.
Situation 2: with the matching rules result is that speed parameter sends to Arduino control desk 33.
Situation 3: many cooresponding speed parameters of rule are identical; Then directly speed parameter is sent to the Arduino control desk, otherwise, according to coverage; Calculate the speed parameter of the maximum cooresponding matching rules of coverage, the speed parameter that calculates is sent to Arduino control desk 33.Wherein ask the maximum formula of coverage to be:
max(|Cov(Cl
t)|)
Cl
tBe decision value, r
jBe O
iMatching rules is concentrated has identical decision value (Cl
tIdentical) rule, cov (r
j) be regular r
jCoverage, Cov (Cl
t) expression has the union of coverage of identical decision attribute value.
Max (| Cov (Cl
t) |) expression coverage most number.
Especially, having under the situation of coupling and the speed parameter of output is 0, promptly miniature intelligent vehicle stops, and it is regular just need not to continue to mate the corner of controlling miniature intelligent vehicle.
For example, detection module 20 detected information are shown in the table 11 when control in real time:
Table 11
Control rate rule in this information and the above-mentioned table 10 is complementary, and can matching rules be 2,4 two rules, and output information is rule 2, rule 4 cooresponding speed parameter and coverages, and is as shown in table 12:
Table 12
Belong to situation 3; At first calculate and have the coverage sum of identical decision value according to formula
; Because the speed parameter in the Rule Information of output is inequality in the case, so Cov (0)={ T5, T6}; Cov (195)=T4}, so max (| Cov (195) |)=2.Therefore be 0 to give the Arduino control desk by controlled variable computing module 32 transmission speed parameters.
Especially, the controlled variable that send this moment is 0, representes that miniature intelligent vehicle stops, and control unit 31B need not continue to mate the rule of controlling miniature intelligent vehicle corner during historical facts or anecdotes; Correspondingly, controlled variable computing module 32 also need not continue to calculate the corner parameter.
If the speed parameter that promptly calculates is not 0; Then need continue to mate the rule of the miniature intelligent vehicle corner of control under the corresponding speed; Wherein when coupling corner rule, to carry out 3 times coupling; The rule of the direction of coupling control earlier, coupling is controlled the rule of angular interval again, the rule of the concrete angle of coupling control at last.Accordingly, the calculation control parameter module needs the parameter of parameter, the parameter between the calculation control corner regions and the concrete angle of calculation control of calculation control corner direction equally.
Regular similar with the coupling control rate, as condition, the information of detection module 20 detections and the rule condition of the control direction among the memory device 31C mate, and export these Rule Informations with the rule of control direction among the memory device 31C.
Likewise, the Rule Information of control unit 31B output in real time mainly is divided into three kinds of situation: (1), the rule that is not complementary, direction parameter are empty, and coverage is empty; (2) have only a rule coupling, a direction parameter, coverage is the cooresponding coverage of matching rules; (3) many rule couplings are arranged, one or more direction parameters, coverage is the cooresponding coverage of each bar control law.
Controlled variable computing module 32, the control law information calculations direction parameter of exporting according to the rule of real-time control unit 31B coupling control direction, and according to circumstances direction parameter is sent to Arduino control desk 33.
Situation 1: the corner of miniature intelligent vehicle is constant, and the corner parameter that controlled variable computing module 32 sent when controlling last time is given Arduino control desk 33, and control unit 31B need not continue to mate the rule between the control corner regions in real time.
Situation 2: obtain the cooresponding corner direction of this matching rules result; If controlled variable is craspedodrome then sends the parameter of keeping straight on by controlled variable computing module 32 and give Arduino control desk (having only a corner parameter under the craspedodrome situation) that real-time control unit 31B need not continue to mate the rule of controlling between corner regions; Otherwise do not send controlled variable and give Arduino control desk 33, control unit 31B need continue to mate the rule between the control corner regions in real time.
Situation 3: calculate max (Cov (Cl
t)) direction of cooresponding corner, if controlled variable for keeping straight on then send the parameter of keeping straight on by controlled variable computing module 32 and give the Arduino control desk, control unit 31B need not continue to mate the rule between the control corner regions in real time; Otherwise do not send controlled variable and give Arduino control desk 33, control unit 31B need continue to mate the rule between the control corner regions in real time.
Method of calculating is with identical with the method for aforementioned calculation speed parameter.
In the rule of above-mentioned coupling control direction, if the direction that controlled variable computing module 32 obtains is not left-hand rotation or right-hand rotation, but keep straight on, then need not continue to mate rule and the rule of controlling concrete angle between the control corner regions.
Likewise, regular similar with the coupling control rate, the rule of controlling between corner regions in control unit 31B and the memory device in real time is complementary.Likewise, the Rule Information of control module 31B output in real time has three kinds of situation: (1), the rule that is not complementary, corner interval parameter are empty, and coverage is empty; (2) have only a rule coupling, a corner interval parameter, coverage is the cooresponding coverage of matching rules; (3) many rule couplings are arranged, one or more corner interval parameters, coverage is the cooresponding coverage of each bar control law.
Controlled variable computing module 32 is controlled the control law information calculations corner interval parameter that the rule between corner regions is exported according to real-time control unit 31B coupling, and is sent the corner parameter according to circumstances and give Arduino control desk 33
Situation 1: if coupling corner direction is turn left (right-hand rotation); Then the cooresponding corner parameter of intermediate value under controlled variable computing module 32 transmission left-hand rotation (right-hand rotation) situation is given Arduino control desk 33, and control unit 31B need not continue to mate the rule between the control corner regions in real time.As: the miniature intelligent vehicle left-hand rotation interval of the present invention's experiment is [81,118], and the right-hand rotation interval is [42,79], and craspedodrome is 80.Then mate the corner direction under the situation of turning left, send the corner parameter and be 99 value and give the Arduino control desk; Under the right-hand rotation situation, send the corner parameter and be 61 value and give the Arduino control desk.
Situation 2: obtain between the cooresponding corner regions of this matching rules result, controlled variable computing module 32 does not send controlled variable and gives Arduino control desk 33, and control unit 31B need continue to mate the rule between the control corner regions in real time.
Situation 3: calculating max (| Cov (Cl
t) |) direction of cooresponding corner, controlled variable computing module 32 does not send controlled variable and gives Arduino control desk 33, and control unit 31B need continue to mate the rule between the control corner regions in real time;
Method of calculating is with identical with the method for aforementioned calculation speed parameter.
Likewise, regular similar with the coupling control rate, the rule of controlling in real time concrete corner in control unit 31B and the memory device is complementary.The Rule Information of same control law module 31 outputs has three kinds of situation, and is identical with above-mentioned three kinds of situation: (1), the rule that is not complementary, and concrete corner parameter is empty, coverage is empty; (2) have only a rule coupling, a concrete corner parameter, coverage are the cooresponding coverage of matching rules; (3) many rule couplings are arranged, one or more concrete corner parameters, coverage are the cooresponding coverage of each bar control law.
Controlled variable computing module 32, the concrete corner parameter of exporting according to the rule of the concrete corner of real-time control unit 31B coupling control of control law information calculations, and this corner parameter sent to Arduino control desk 33
Situation 1: between the coupling corner regions is [l, r], then, as concrete angle parameter, (l+r)/2 is sent to Arduino control desk 33 with (l+r)/2.
Situation 2: controlled variable computing module 32 sends to Arduino control desk 33 with the cooresponding corner parameter of matching rules.
Situation 3: calculating max (| Cov (Cl
t) |) cooresponding corner parameter, controlled variable computing module 32 transmission max (| Cov (Cl
t) |) cooresponding corner parameter gives Arduino control desk 33.
Method of calculating is with identical with the method for aforementioned calculation speed parameter.
Said Arduino control desk 33 converts speed parameter to motor control signal, sends to motor driver 40, becomes the steering wheel control signal to send to steering wheel actuator 50 the corner Parameters Transformation;
Said Arduino control desk is an I/O interface control desk that comes based on open true form development; Has the development environment that uses JAVA, C language; But let user's at high speed use Arduino language and softwares such as FLASH or Processing, make more how interactive product.The electronic component that Arduino can use other exploitation to accomplish is like output units such as LED, stepping motors.Simultaneously, Arduino also can independently operate and be called the platform that can follow software handshake.
The miniature intelligent vehicle intelligence control method of the present invention, preferred embodiment is as shown in Figure 5, comprising:
201, utilize camera collection road shape signal;
202, from the road conditions signal, detect road condition information, comprise any one or multiple combination arbitrarily in obstacle information, traffic lights information, taper mark information, lane mark information, the traffic sign information;
203, mate road condition information and control law, export the control law information of all couplings, comprise the result and the cooresponding coverage of the control law of coupling;
Said control law obtains through the mode of step 101-104 under off-line state, repeats no more.
Said matching way is seen the description of real-time control unit 31B, repeats no more.
204, according to the control law information calculations speed parameter and the corner parameter of all couplings, speed parameter and corner parameter are sent to the Arduino control desk;
Specifically, repeat no more referring to the description of controlled variable computing module 32.
205, the Arduino control desk converts speed parameter to motor control signal respectively, and the corner Parameters Transformation is become the steering wheel control signal.
Main task of the present invention is to use the Based Intelligent Control that the grain theory of computation realizes miniature intelligent vehicle; The precise math model of having avoided traditional control to set up; Controlling through setting up the control law of level in real time; Progressively add smart control method, have the better controlling precision, real-time is good and antijamming capability is strong.
Embodiment that the present invention lifts or embodiment have carried out further detailed description to the object of the invention, technical scheme and advantage; Institute is understood that; Abovely lift embodiment or embodiment is merely preferred implementation of the present invention; Not in order to restriction the present invention, all within spirit of the present invention and principle to any modification that the present invention did, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (10)
1. miniature intelligent vehicle intelligence controlling device; Comprise camera (10), signal detection module (20), control law module (31), controlled variable computing module (32), Arduino control desk (33), motor driver (40) and steering wheel actuator (50); It is characterized in that
Said signal detection module (20) detects road condition information from the information that camera (10) transmits; Said road condition information is any one the perhaps multiple combination arbitrarily in obstacle information, traffic lights information, taper mark information, lane mark information, the traffic sign information;
Said control law module (31) comprises the control law acquiring unit (31A) and the real-time control unit (31B) that are connected through memory device (31C);
Said control law acquiring unit (31A) is used to obtain control law, and control law is stored in memory device (31C), and said control law calculates acquisition according to road condition information and current speed parameter and the corner parameter of miniature intelligent vehicle through grain;
Said real-time control unit (31B); Be used to all control law information of obtaining to mate; The result of the control law that promptly is complementary and cooresponding coverage; All control law information of output coupling are given controlled variable computing module (32), and all control law information of said coupling are mated according to the road condition information of obtaining from signal detection module (20) with from the control law that memory device (31C) obtains and obtained;
Said controlled variable computing module (32), control law information calculations speed parameter and corner parameter according to real-time control unit (31B) output send to Arduino control desk (33) with speed parameter and corner parameter;
Said Arduino control desk (33) converts speed parameter to motor control signal, and motor control signal is sent to motor driver (40), and the corner Parameters Transformation is become the steering wheel control signal, and the steering wheel control signal is sent to steering wheel actuator (50).
2. according to the said miniature intelligent vehicle intelligence controlling device of claim 1; It is characterized in that said signal detection module (20) comprises any one or any multiple combination in detection of obstacles unit (21), traffic lights detecting unit (22), taper mark detecting unit (23), lane mark detecting unit (24) and the traffic sign detecting unit (25); Said detection of obstacles unit (21), traffic lights detecting unit (22), taper mark detecting unit (23), lane mark detecting unit (24) and traffic sign detecting unit (25) detect obstacle information, traffic lights information, taper mark information, lane mark information and traffic sign information respectively from the information that camera (10) transmits.
3. according to the said miniature intelligent vehicle intelligence controlling device of claim 1; It is characterized in that; The said control law of control law acquiring unit (31A) calculates acquisition according to road condition information and current speed parameter and the corner parameter of miniature intelligent vehicle through grain, comprising:
(1) the grain definition of miniature intelligent vehicle and the calculating of grain;
(2) use progressively decomposed particles of interparticle operation rule, at first, obtain the rule of control rate; , speed obtains the rule of control corner then under not being 0 situation;
The rule of wherein obtaining the control corner comprises: obtain the rule of control direction, obtain the rule between the control corner regions, obtain the rule of the concrete angle of control.
4. according to the said miniature intelligent vehicle intelligence controlling device of claim 1; It is characterized in that; In real time all control law information of the said coupling of control unit (31B) are mated according to the road condition information of obtaining from signal detection module (20) with from the control law that memory device (31C) obtains and are obtained, and comprising:
If the detected road condition information of signal detection module (20) satisfies the condition of control law in the memory device (31C), then export this or many control law information, promptly should the cooresponding speed parameter of rule or corner parameter and this regular coverage;
Wherein the control law information of output comprises:
(1) rule that is not complementary, speed parameter or corner parameter are empty, and coverage is empty;
(2) have only a matching rules, a speed parameter or corner parameter are only arranged, coverage is the cooresponding coverage of matching rules;
(3) many matching rules are arranged, one or more speed parameters or corner parameter are arranged, coverage is the cooresponding coverage of each bar control law.
5. according to claim 1 or 4 said miniature intelligent vehicle intelligence controlling devices; It is characterized in that; The control law information calculations speed parameter and the corner parameter of the real-time control unit of the said basis of controlled variable computing module (32) (31B) output, and speed parameter and corner parameter sent to Arduino control desk (33);
Said computation speed parameter comprises:
(1), the speed parameter of last time is sent to Arduino control desk (33) if the rule that is not complementary;
(2), the speed parameter of matching rules is sent to Arduino control desk (33) if having only a matching rules;
(3) if many matching rules are arranged; If many cooresponding speed parameters of rule are identical; Then directly speed parameter is sent to the Arduino control desk, otherwise, according to coverage; Calculate the speed parameter of the maximum cooresponding matching rules of coverage, the speed parameter that calculates is sent to Arduino control desk (33);
Said calculating corner parameter comprises: at first, be not under 0 the situation in speed, and the calculated direction parameter; Secondly, not under the parameter situation of keeping straight at direction parameter, calculate the parameter between corner regions; At last, calculate concrete corner parameter.
6. miniature intelligent vehicle intelligence control method comprises:
11, utilize camera collection road conditions signal;
12, from the road conditions signal, detect road condition information, comprise any one or multiple combination arbitrarily in obstacle information, traffic lights information, taper mark information, lane mark information, the traffic sign information;
13, coupling road condition information and control law are exported the control law information that all are complementary, the result of the control law that promptly is complementary and cooresponding coverage;
14, according to the control law information calculations speed parameter and the corner parameter of all couplings, speed parameter and corner parameter are sent to the Arduino control desk;
15, the Arduino control desk converts speed parameter to motor control signal respectively, and the corner Parameters Transformation is become the steering wheel control signal.
7. according to the said miniature intelligent vehicle intelligence control method of claim 6, it is characterized in that said control law calculates acquisition according to road condition information and current speed parameter and the corner parameter of miniature intelligent vehicle through grain, comprising:
(1) the grain definition of miniature intelligent vehicle and the calculating of grain
(2) use progressively decomposed particles of interparticle operation rule, at first, obtain the rule of control rate; , speed obtains the rule of control corner then under not being 0 situation;
The rule of wherein obtaining the control corner comprises: obtain the rule of control direction, obtain the rule between the control corner regions, obtain the rule of the concrete angle of control.
8. according to the said miniature intelligent vehicle intelligence control method of claim 6, it is characterized in that said coupling road condition information and control law are exported the control law information of all couplings, comprising:
If detected road condition information satisfies the condition of control law in the memory device, then export one or more control law information, promptly should the cooresponding speed parameter of rule or corner parameter and this regular coverage;
Wherein the control law information of output comprises:
(1) rule that is not complementary, speed parameter or corner parameter are empty, and coverage is empty;
(2) have only a matching rules, a speed parameter or corner parameter are only arranged, coverage is the cooresponding coverage of matching rules;
(3) many matching rules are arranged, one or more speed parameters or corner parameter are arranged, coverage is the cooresponding coverage of each bar control law.
9. according to claim 6 or 8 said miniature intelligent vehicle intelligence control methods, it is characterized in that said control law information calculations speed parameter and corner parameter according to all couplings send to the Arduino control desk with speed parameter and corner parameter;
Said computation speed parameter comprises:
(1), the speed parameter of last time is sent to the Arduino control desk if the rule that is not complementary;
(2), the speed parameter of matching rules is sent to the Arduino control desk if having only a matching rules;
(3) if many matching rules are arranged; If many cooresponding speed parameters of rule are identical; Then directly speed parameter is sent to the Arduino control desk, otherwise, according to coverage; Calculate the speed parameter of the maximum cooresponding matching rules of coverage, the speed parameter that calculates is sent to the Arduino control desk;
Said calculating corner parameter comprises: at first, be not under 0 the situation in speed, and the calculated direction parameter; Secondly, not under the parameter situation of keeping straight at direction parameter, calculate the parameter between corner regions; At last, calculate concrete corner parameter.
10. according to the said miniature intelligent vehicle intelligence control method of claim 9, it is characterized in that, when speed parameter is 0, no longer calculate the corner parameter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210249480.6A CN102745196B (en) | 2012-07-18 | 2012-07-18 | Intelligent control device and method for granular computing-based micro intelligent vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210249480.6A CN102745196B (en) | 2012-07-18 | 2012-07-18 | Intelligent control device and method for granular computing-based micro intelligent vehicle |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102745196A true CN102745196A (en) | 2012-10-24 |
| CN102745196B CN102745196B (en) | 2015-02-18 |
Family
ID=47025808
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210249480.6A Active CN102745196B (en) | 2012-07-18 | 2012-07-18 | Intelligent control device and method for granular computing-based micro intelligent vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102745196B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104015723A (en) * | 2014-06-12 | 2014-09-03 | 北京工业大学 | Intelligent vehicle control system and method based on intelligent transportation platform |
| CN106297522A (en) * | 2016-10-31 | 2017-01-04 | 中国科学院自动化研究所 | A kind of miniature sand table demonstration system and method |
| CN107316486A (en) * | 2017-07-11 | 2017-11-03 | 湖南星云智能科技有限公司 | Pilotless automobile visual identifying system based on dual camera |
| CN107609472A (en) * | 2017-08-04 | 2018-01-19 | 湖南星云智能科技有限公司 | A kind of pilotless automobile NI Vision Builder for Automated Inspection based on vehicle-mounted dual camera |
| WO2019047644A1 (en) * | 2017-09-05 | 2019-03-14 | 百度在线网络技术（北京）有限公司 | Method and device for controlling autonomous vehicle |
| WO2022048164A1 (en) * | 2020-09-03 | 2022-03-10 | 哈尔滨工业大学 | Smart robot dialogue policy generation method based on granular computing |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004518073A (en) * | 2001-02-07 | 2004-06-17 | シーメンス アクチエンゲゼルシヤフト | Control method of automobile power train |
| JP2009053909A (en) * | 2007-08-27 | 2009-03-12 | Fuji Xerox Co Ltd | Self-propelled device and self-propelled control program |
| CN101436037A (en) * | 2008-11-28 | 2009-05-20 | 深圳先进技术研究院 | Dining room service robot system |
| CN101734252A (en) * | 2009-12-23 | 2010-06-16 | 合肥工业大学 | Preview tracking control unit for intelligent vehicle vision navigation |
| CN102541061A (en) * | 2012-02-07 | 2012-07-04 | 清华大学 | Micro intelligent vehicle based on visual and auditory information |
| CN102541063A (en) * | 2012-03-26 | 2012-07-04 | 重庆邮电大学 | Line tracking control method and line tracking control device for micro intelligent automobiles |
-
2012
- 2012-07-18 CN CN201210249480.6A patent/CN102745196B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004518073A (en) * | 2001-02-07 | 2004-06-17 | シーメンス アクチエンゲゼルシヤフト | Control method of automobile power train |
| JP2009053909A (en) * | 2007-08-27 | 2009-03-12 | Fuji Xerox Co Ltd | Self-propelled device and self-propelled control program |
| CN101436037A (en) * | 2008-11-28 | 2009-05-20 | 深圳先进技术研究院 | Dining room service robot system |
| CN101734252A (en) * | 2009-12-23 | 2010-06-16 | 合肥工业大学 | Preview tracking control unit for intelligent vehicle vision navigation |
| CN102541061A (en) * | 2012-02-07 | 2012-07-04 | 清华大学 | Micro intelligent vehicle based on visual and auditory information |
| CN102541063A (en) * | 2012-03-26 | 2012-07-04 | 重庆邮电大学 | Line tracking control method and line tracking control device for micro intelligent automobiles |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104015723A (en) * | 2014-06-12 | 2014-09-03 | 北京工业大学 | Intelligent vehicle control system and method based on intelligent transportation platform |
| CN104015723B (en) * | 2014-06-12 | 2016-08-24 | 北京工业大学 | A kind of intelligent vehicle control system and method based on intelligent transportation platform |
| CN106297522A (en) * | 2016-10-31 | 2017-01-04 | 中国科学院自动化研究所 | A kind of miniature sand table demonstration system and method |
| CN107316486A (en) * | 2017-07-11 | 2017-11-03 | 湖南星云智能科技有限公司 | Pilotless automobile visual identifying system based on dual camera |
| CN107609472A (en) * | 2017-08-04 | 2018-01-19 | 湖南星云智能科技有限公司 | A kind of pilotless automobile NI Vision Builder for Automated Inspection based on vehicle-mounted dual camera |
| WO2019047644A1 (en) * | 2017-09-05 | 2019-03-14 | 百度在线网络技术（北京）有限公司 | Method and device for controlling autonomous vehicle |
| WO2022048164A1 (en) * | 2020-09-03 | 2022-03-10 | 哈尔滨工业大学 | Smart robot dialogue policy generation method based on granular computing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102745196B (en) | 2015-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Prakash et al. | Multi-modal fusion transformer for end-to-end autonomous driving | |
| US10803328B1 (en) | Semantic and instance segmentation | |
| US20210397185A1 (en) | Object Motion Prediction and Autonomous Vehicle Control | |
| US20190145765A1 (en) | Three Dimensional Object Detection | |
| WO2020135810A1 (en) | Multi-sensor data fusion method and device | |
| CN102745196B (en) | Intelligent control device and method for granular computing-based micro intelligent vehicle | |
| Cai et al. | Vision-based trajectory planning via imitation learning for autonomous vehicles | |
| Bucher et al. | Image processing and behavior planning for intelligent vehicles | |
| CN103454919B (en) | The control method of the kinetic control system of mobile robot in intelligent space | |
| Hecker et al. | Learning accurate, comfortable and human-like driving | |
| CN116323364A (en) | Waypoint prediction and motion forecast for vehicle motion planning | |
| US12013457B2 (en) | Systems and methods for integrating radar data for improved object detection in autonomous vehicles | |
| US11975742B2 (en) | Trajectory consistency measurement for autonomous vehicle operation | |
| US20230111354A1 (en) | Method and system for determining a mover model for motion forecasting in autonomous vehicle control | |
| CN112130570A (en) | Blind guiding robot of optimal output feedback controller based on reinforcement learning | |
| Wang et al. | End-to-end driving simulation via angle branched network | |
| Cao et al. | A learning-based vehicle trajectory-tracking approach for autonomous vehicles with lidar failure under various lighting conditions | |
| CN115635961A (en) | Sample data generation method and trajectory prediction method and device applying same | |
| Dong et al. | A vision-based method for improving the safety of self-driving | |
| John et al. | Estimation of steering angle and collision avoidance for automated driving using deep mixture of experts | |
| Souza et al. | Vision-based waypoint following using templates and artificial neural networks | |
| CN105867112A (en) | Intelligent vehicle based on control algorithm with automatically optimized parameter and control method thereof | |
| Wang et al. | Vision‐Based Lane Departure Detection Using a Stacked Sparse Autoencoder | |
| CN110297489A (en) | The control method and control system of automatic driving vehicle | |
| US20210142225A1 (en) | Ensemble of narrow ai agents |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |