CN107943066A - It is a kind of to have the man-machine method for supervision and control to unmanned plane obstacle avoidance - Google Patents
It is a kind of to have the man-machine method for supervision and control to unmanned plane obstacle avoidance Download PDFInfo
- Publication number
- CN107943066A CN107943066A CN201710553624.XA CN201710553624A CN107943066A CN 107943066 A CN107943066 A CN 107943066A CN 201710553624 A CN201710553624 A CN 201710553624A CN 107943066 A CN107943066 A CN 107943066A
- Authority
- CN
- China
- Prior art keywords
- unmanned plane
- action
- obstacle avoidance
- obstacle
- supervised control
- 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
- 238000000034 method Methods 0.000 title claims abstract description 28
- 230000009471 action Effects 0.000 claims abstract description 152
- 238000004891 communication Methods 0.000 claims abstract description 35
- 230000007613 environmental effect Effects 0.000 claims abstract description 30
- 238000011084 recovery Methods 0.000 claims abstract description 3
- 230000004888 barrier function Effects 0.000 claims description 33
- 230000000875 corresponding effect Effects 0.000 claims description 19
- 230000008569 process Effects 0.000 claims description 11
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 9
- 230000002452 interceptive effect Effects 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000003993 interaction Effects 0.000 claims description 4
- 235000013399 edible fruits Nutrition 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 2
- 238000007726 management method Methods 0.000 description 27
- 108091032995 L1Base Proteins 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000465 moulding Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000004886 process control Methods 0.000 description 1
- 230000008929 regeneration Effects 0.000 description 1
- 238000011069 regeneration method Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
There is the man-machine method for supervision and control to unmanned plane obstacle avoidance the present invention provides a kind of, it is related to unmanned plane field, the present invention defines the value of environmental factor, obstacle avoidance action and Supervised Control pattern, unmanned plane makes inferences judgement to the action of optional obstacle avoidance, judged according to judging result, so that obstacle avoidance, the present invention is for known obstacle, unmanned plane gives full play to its capacity of will, complete independently obstacle avoidance;For unknown obstacle, unmanned plane must be completed to evade in the case where there is man-machine guidance, and variable autonomous method for supervision and control gives full play to the autonomous analysis judgement performed and have man-machine operator of unmanned plane;For extreme cases such as communication disruptions, unmanned plane passes through the adjusting to Supervised Control pattern, ensure the safety of itself while communication recovery is waited, variable autonomous Supervised Control pattern combination unmanned plane with have the characteristics of man-machine, have preferable adaptability to changes to different the barrer types and environmental aspect.
Description
Technical field
The present invention relates to unmanned plane field, especially one kind to evade determination methods.
Background technology
When facing barrier in unmanned plane traveling process, there is man-machine operator to be interacted with unmanned plane, ensure that unmanned plane is complete
Into obstacle avoidance.Supervised Control refer to have it is man-machine intermittently interacted with unmanned plane, receive and feed back and provide instruction with right
Unmanned plane in task environment carries out process control.
However, in the practical application of UAV system obstacle avoidance, traditional method is generally grasped manually using operator
Make carry out obstacle avoidance or carry out obstacle rule by way of calculating cost function and using path planning, rule or automatic machine
Keep away.Two problems so occur:1. " excess load " phenomenon of operator is caused due to the complexity of environment;2. due to nothing
Man-machine system is in excessive autonomous authority, cause operator lose to surrounding environment situational awareness " people not return
Road " (Out-of-the-Loop, OOTL) phenomenon.Two problem explanations, there is the man-machine Supervised Control to unmanned plane obstacle avoidance
Need in the light of actual conditions dynamic regulation.
Therefore, be highly desirable research one kind can either give full play to the high-level cognitive Decision ability of man-machine operator and
Unmanned plane efficiently performs task ability, while maintains human-machine operation person to the situation cognition of environment and appropriate live load
Unmanned plane obstacle avoidance method for supervision and control.
The content of the invention
For overcome the deficiencies in the prior art, the present invention proposes a kind of obstacle avoidance method for supervision and control, can cause people
Machine has complementary advantages, while can keep perception and appropriate live load of the operator to environmental situation again, avoids Supervised Control
Middle two problems occurred, so as to effectively realize the purpose of obstacle avoidance.
Obstacle avoidance method for supervision and control proposed by the present invention, is a kind of according to environmental change, there is the different prisons of man-machine adjustment
Control model is superintended and directed, unmanned plane completes corresponding obstacle avoidance action, to realize the method for obstacle avoidance purpose.
The technical solution adopted by the present invention to solve the technical problems comprises the following steps:
Step1:Define environmental factor, obstacle avoidance action and Supervised Control pattern
Unmanned plane has 4 kinds of obstacle avoidance actions, including autonomous obstacle avoidance action A1, deceleration etc. when running into barrier
Treat instruction action A2, spiraling waits instruction action A3 and return action A4;
Wherein, it is high-grade action to define autonomous obstacle avoidance action A1;Slow down and wait instruction A2 to refer to wait of spiraling
Make the action that A3 is middle grade;Return to the action that A4 is inferior grade;
Supervised Control pattern is 3 patterns, and the capacity of will of unmanned plane is respectively management by exception pattern L3 from high to low, same
Anticipate management mode L2 and instruction control model L1;Outer management mode L3 exemplified by the initial pattern of unmanned plane;
It is as shown in table 1 to define unmanned plane Supervised Control pattern;
1 three kinds of Supervised Control patterns of table and its respective action
Action executing mode refers to unmanned plane in the case where there is man-machine Supervised Control, completes the side of corresponding obstacle avoidance action
Formula, as shown in table 2:
2 three kinds of Supervised Control patterns of table and its corresponding action executing mode
1) environmental factor includes obstacle distance distance, whether obstacle information is known and whether current communication status is good
3 aspects, are defined as follows:
1. the definition of obstacle distance, including the nearly V3 of the distance of obstacle and remote V5 of distance of obstacle;
Relative to unmanned plane, the position of barrier can be expressed as
In formula (1), P (t) is the position of t moment barrier i, and t is the time, and P (0) is the initial position of barrier i, viFor
The speed of barrier i, vu(τ) is the speed of unmanned plane u;
In formula (2)For the yaw angle of unmanned plane, the distance of unmanned plane and obstacle distance is by arrival Obstacle Position
Time T defined by formula below:
2. whether barrier is known to be divided into known obstacle V1 and unknown obstacle V4, it is known that obstacle refers to the obstacle in operating area
The image information and positional information of thing are it is known that and unknown obstacle is the image information of barrier and position in operating area
At least one information of information is unknown;
3. current communication status is divided into communication normally and interrupts V2, the signal that communication disruption is pointed to up to unmanned plane receiver is strong
Degree, which is less than, the sensitivity of man-machine receiver, cause to have it is man-machine between unmanned plane can not normal communication situation;
2) codomain of environmental factor and the value of obstacle avoidance action is { 0,1 }, wherein 1 represents that the state is effective, 0 table
Show that state is invalid;Specific environment factor and the value of operating state such as table 3 below, shown in table 4:
The value of 3 environmental factor of table
Wherein * represents outlier, and outlier represents that value can be " 0 " or " 1 ", and the value does not influence corresponding shape under *
State, only influencing it needs to take the state under particular value;
The value of 4 unmanned plane operating state of table
Step2:The reasoning and judging of the optional obstacle avoidance action of unmanned plane
1) state vector is write out by Step1
State vector represents that wherein k represents number with C (k), and C (0) represents initial state vector, and state vector includes
Obstacle avoidance acts and the full content of environmental factor, form are:
Wherein, during no avoiding action, the value of acquiescence unmanned plane A1~A4 is all 0, i.e.,
V1~V5 is the value that Step1 is obtained, and is always definite value in Step2;
2) made inferences using state vector C (k), reasoning process is as follows:
1. multiplying adjacent weight matrix W with state vector C (k) obtains intermediate vector X (k):
X (k)=C (k) W (6)
Wherein, the adjoining weight matrix W between node is as follows:
2. to intermediate vector X (k) each component x of state transition function f (x) processing intermediate vector X (k), state turns
It is two-value jump function to move function f (x):
Wherein, x is the component of intermediate vector X (k), and vectorial X (k) dimension is 6, and the codomain of component x is { 0,1 };
3. state vector C (k) is updated with intermediate vector X (k), i.e.,
4. continuous repeat step 1., step 2. with step 3., that is, constantly update state vector C (k), until state to
Measure C (k+n+1)=C (k+n), wherein, n represents number, i.e., the value of (k+n+1) next state vector and (k+n) next state to
The value of amount is identical;
C (k+n+1)=C (k+n)=[a1, a2, a3, a4 | V1, V2, V3, V4, V5], wherein a1, a2, a3, a4 difference table
Show that obstacle avoidance acts A1, A2, A3, the corresponding values of A4, are worth for 0 or 1, and value is 1 item in a1, a2, a3, a4, corresponding
Obstacle avoidance moves to be acted for optional obstacle avoidance;
It is man-machine to having that 5. unmanned plane provides optional obstacle avoidance the result of the action;
Step3:Final obstacle avoidance action judges
1) when meeting following condition a) and b) wherein one, then unmanned plane is needed with there is man-machine interact:
A) be " 1 " comprising unknown obstacle V4 entry value in environmental factor, i.e. V=(*, *, *, 1, *), wherein, * represents unrelated
, outlier represents that value is " 0 " or " 1 ";
B) the current Supervised Control pattern respective action grade of unmanned plane is closed in optional obstacle avoidance performance of a different dive;
Action executing mode in current Supervised Control pattern is as follows:
1. when the Supervised Control pattern of unmanned plane is management by exception pattern L3, and unmanned plane needs interactive, as in 15s nobody
The obstacle avoidance action reasoning that machine can not provide is as a result, i.e. barrier is unknown, then unmanned plane reduces Supervised Control pattern to agreement
Management mode L2;If unmanned plane is to having man-machine offer obstacle avoidance action reasoning as a result, unmanned plane is using intelligent combination master
Dynamic feedback, operator do not negate the obstacle avoidance action reasoning for performing unmanned plane offer in 15s as a result, conversely, operator
Negate not perform in 15s, and reduce Supervised Control pattern to agreement management mode L2;
2. when the Supervised Control pattern of unmanned plane is agrees to management mode L2, when unmanned plane needs interactive, as in 15s nobody
The obstacle avoidance action reasoning that machine can not provide is as a result, i.e. barrier is unknown, then unmanned plane reduces Supervised Control pattern to agreement
Management mode L1;If unmanned plane is to having man-machine offer obstacle avoidance action reasoning as a result, unmanned plane is using intelligent combination master
Dynamic feedback, waits the accreditation of operator, and operator does not negate to perform the obstacle avoidance action that unmanned plane provides to push away in 15s
Reason is not as a result, no negative performs then in 15s, and reduces Supervised Control pattern to operator's decision-making mode L1;
3. when the Supervised Control pattern of unmanned plane is operator decision-making mode L1, and unmanned plane needs interactive, unmanned plane to
The image information and positional information of man-machine offer barrier are provided, and state vector C (k) and obstacle avoidance action reasoning are provided
As a result, there is man-machine combined using pipe-connecting mode to perform:
If (known to the barrer types), then (select autonomous obstacle avoidance decision-making to evade)
If (the barrer types are unknown and apart from remote), then (selection, which is slowed down, waits instruction)
If (the barrer types are unknown and apart near), then (selection, which is spiraled, waits instruction) (9)
Unmanned plane obstacle avoidance is provided by order to act, and the decision-making of operator is waited in unmanned plane 15s, according to decision-making knot
Fruit performs, and time-out then comes back to the base without decision-making;
2) when the action of optional obstacle avoidance meets the current Supervised Control pattern respective action grade of unmanned plane, and environment because
When element does not occur the situation of Rule of judgment a) of step 1) in step2, unmanned plane is not required to and has human-computer interaction, optional obstacle rule
Action is kept away to perform according to following priority:
A1>A2>A3>A4 (10)
Action executing mode is as follows in corresponding Supervised Control pattern:
1. unmanned plane need not interact, when the action under management by exception pattern L3, then allowing to perform is autonomous obstacle
Avoiding action A1, wait instruction action of slowing down A2, spiraling waits instruction action A3 and return action A4, and unmanned plane is according to formula
(10) priority uses autonomous executive mode;
2. unmanned plane need not interact, when in the case where agreeing to management mode L2, it is allowed to which the action of execution refers to for deceleration wait
Order action A2, spiraling waits instruction action A3 and return action A4, and unmanned plane is according to formula (10) priority using the autonomous side of execution
Formula;
3. unmanned plane need not interact, when in the case where instructing control model L1, it is allowed to which the action of execution is return action A4;
Unmanned plane uses autonomous executive mode according to formula (10) priority;
Step4:The environmental factor of Step1 of renewal and the value of obstacle avoidance action, unmanned plane Supervised Control mould per 30s
If formula reduces, renewal unmanned plane Supervised Control pattern is the pattern after reduction, is moved in circles according to Step1~Step3, nobody
Machine performs the obstacle avoidance finally obtained and acts, completion obstacle avoidance, after success obstacle avoidance, unmanned plane Supervised Control pattern
Revert to management by exception pattern L3.
The beneficial effects of the invention are as follows can give full play to its capacity of will, complete independently for known obstacle, unmanned plane
Obstacle avoidance;For unknown obstacle, unmanned plane must be completed to evade in the case where there is man-machine guidance, variable autonomous Supervised Control side
Method can give full play to the autonomous analysis judgement performed and have man-machine operator of unmanned plane;For poles such as communication disruptions
End situation, unmanned plane can ensure the peace of itself by the adjusting to Supervised Control pattern while communication recovery is waited
Entirely.Variable autonomous Supervised Control pattern can combine unmanned plane and have the characteristics of man-machine, to different the barrer types and environment shape
Condition has preferable adaptability to changes.
Brief description of the drawings
Fig. 1 is the model framework chart of unmanned plane obstacle avoidance of the present invention.
Fig. 2 is the obstacle avoidance procedure chart of the embodiment of the present invention 1.
Fig. 3 is the obstacle avoidance procedure chart of the embodiment of the present invention 2.
Fig. 4 is the obstacle avoidance procedure chart of the embodiment of the present invention 3.
Embodiment
The present invention is further described with reference to the accompanying drawings and examples, and Fig. 1 is the mould of unmanned plane obstacle avoidance of the present invention
Type block diagram.
Step1:Define environmental factor, obstacle avoidance action and Supervised Control pattern
Unmanned plane has 4 kinds of obstacle avoidance actions, including autonomous obstacle avoidance action A1, wait of slowing down when running into barrier
Instruction action A2, spiraling waits instruction action A3 and return action A4;
Wherein, it is high-grade action to define autonomous obstacle avoidance action A1;Slow down and wait instruction A2 to refer to wait of spiraling
Make the action that A3 is middle grade;Return to the action that A4 is inferior grade;
Supervised Control pattern is 3 patterns, and the capacity of will of unmanned plane is respectively management by exception pattern L3 from high to low, same
Anticipate management mode L2 and instruction control model L1;Outer management mode L3 exemplified by the initial pattern of unmanned plane;
It is as shown in table 1 to define unmanned plane Supervised Control pattern;
1 three kinds of Supervised Control patterns of table and its respective action
Action executing mode refers to unmanned plane in the case where there is man-machine Supervised Control, completes the side of corresponding obstacle avoidance action
Formula, as shown in table 2:
2 three kinds of Supervised Control patterns of table and its corresponding action executing mode
1) environmental factor includes obstacle distance distance, whether obstacle information is known and whether current communication status is good
3 aspects, are defined as follows:
1. the definition of obstacle distance, including the nearly V3 of the distance of obstacle and remote V5 of distance of obstacle;
Relative to unmanned plane, the position of barrier can be expressed as
In formula (1), P (t) is the position of t moment barrier i, and t is the time, and P (0) is the initial position of barrier i, viFor
The speed of barrier i, vu(τ) is the speed of unmanned plane u;
In formula (2)For the yaw angle of unmanned plane, the distance of unmanned plane and obstacle distance is by arrival Obstacle Position
Time T defined by formula below:
2. whether barrier is known to be divided into known obstacle V1 and unknown obstacle V4, it is known that obstacle refers to the obstacle in operating area
The image information and positional information of thing are it is known that and unknown obstacle is the image information of barrier and position in operating area
At least one information of information is unknown;
3. current communication status is divided into communication normally and interrupts V2, the signal that communication disruption is pointed to up to unmanned plane receiver is strong
Degree, which is less than, the sensitivity of man-machine receiver, cause to have it is man-machine between unmanned plane can not normal communication situation;
2) codomain of environmental factor and the value of obstacle avoidance action is { 0,1 }, wherein 1 represents that the state is effective, 0 table
Show that state is invalid;Specific environment factor and the value of operating state such as table 3 below, shown in table 4:
The value of 3 environmental factor of table
Wherein * represents outlier, and outlier represents that value can be " 0 " or " 1 ", and the value does not influence corresponding shape under *
State, only influencing it needs to take the state under particular value;
The value of 4 unmanned plane operating state of table
Step2:The reasoning and judging of the optional obstacle avoidance action of unmanned plane
1) state vector is write out by Step1
State vector represents that wherein k represents number with C (k), and C (0) represents initial state vector, and state vector includes
Obstacle avoidance acts and the full content of environmental factor, form are:
Wherein, during no avoiding action, the value of acquiescence unmanned plane A1~A4 is all 0, i.e.,
V1~V5 is the value that Step1 is obtained, and is always definite value in Step2;
2) made inferences using state vector C (k), reasoning process is as follows:
1. multiplying adjacent weight matrix W with state vector C (k) obtains intermediate vector X (k):
X (k)=C (k) W (6)
Wherein, the adjoining weight matrix W between node is as follows:
2. to intermediate vector X (k) each component x of state transition function f (x) processing intermediate vector X (k), state turns
It is two-value jump function to move function f (x):
Wherein, x is the component of intermediate vector X (k), and vectorial X (k) dimension is 6, and the codomain of component x is { 0,1 };
3. state vector C (k) is updated with intermediate vector X (k), i.e.,
4. continuous repeat step 1., step 2. with step 3., that is, constantly update state vector C (k), until state to
Measure C (k+n+1)=C (k+n), wherein, n represents number, i.e., the value of (k+n+1) next state vector and (k+n) next state to
The value of amount is identical;
C (k+n+1)=C (k+n)=[a1, a2, a3, a4 | V1, V2, V3, V4, V5], wherein a1, a2, a3, a4 difference table
Show that obstacle avoidance acts A1, A2, A3, the corresponding values of A4, are worth for 0 or 1, and value is 1 item in a1, a2, a3, a4, corresponding
Obstacle avoidance moves to be acted for optional obstacle avoidance.
It is man-machine to having that 5. unmanned plane provides optional obstacle avoidance the result of the action;
Step3:Final obstacle avoidance action judges
1) when meeting following condition a) and b) wherein one, then unmanned plane is needed with there is man-machine interact:
A) be " 1 " comprising unknown obstacle V4 entry value in environmental factor, i.e. V=(*, *, *, 1, *), wherein, * represents unrelated
, outlier represents that value is " 0 " or " 1 ";
B) the current Supervised Control pattern respective action grade of unmanned plane is closed in optional obstacle avoidance performance of a different dive;
Action executing mode in current Supervised Control pattern is as follows:
1. when the Supervised Control pattern of unmanned plane is management by exception pattern L3, and unmanned plane needs interactive, as in 15s nobody
The obstacle avoidance action reasoning that machine can not provide is as a result, i.e. barrier is unknown, then unmanned plane reduces Supervised Control pattern to agreement
Management mode L2;If unmanned plane is to having man-machine offer obstacle avoidance action reasoning as a result, unmanned plane is using intelligent combination master
Dynamic feedback, operator do not negate the obstacle avoidance action reasoning for performing unmanned plane offer in 15s as a result, conversely, operator
Negate not perform in 15s, and reduce Supervised Control pattern to agreement management mode L2;
2. when the Supervised Control pattern of unmanned plane is agrees to management mode L2, when unmanned plane needs interactive, as in 15s nobody
The obstacle avoidance action reasoning that machine can not provide is as a result, i.e. barrier is unknown, then unmanned plane reduces Supervised Control pattern to agreement
Management mode L1;If unmanned plane is to having man-machine offer obstacle avoidance action reasoning as a result, unmanned plane is using intelligent combination master
Dynamic feedback, waits the accreditation of operator, and operator does not negate to perform the obstacle avoidance action that unmanned plane provides to push away in 15s
Reason is not as a result, no negative performs then in 15s, and reduces Supervised Control pattern to operator's decision-making mode L1;
3. when the Supervised Control pattern of unmanned plane is operator decision-making mode L1, and unmanned plane needs interactive, unmanned plane to
The image information and positional information of man-machine offer barrier are provided, and state vector C (k) and obstacle avoidance action reasoning are provided
As a result, there is man-machine combined using pipe-connecting mode to perform:
If (known to the barrer types), then (select autonomous obstacle avoidance decision-making to evade)
If (the barrer types are unknown and apart from remote), then (selection, which is slowed down, waits instruction)
If (the barrer types are unknown and apart near), then (selection, which is spiraled, waits instruction) (9)
Unmanned plane obstacle avoidance is provided by order to act, and the decision-making of operator is waited in unmanned plane 15s, according to decision-making knot
Fruit performs, and time-out then comes back to the base without decision-making;
2) when the action of optional obstacle avoidance meets the current Supervised Control pattern respective action grade of unmanned plane, and environment because
When element does not occur the situation of Rule of judgment a) of step 1) in step2, unmanned plane is not required to and has human-computer interaction, optional obstacle rule
Action is kept away to perform according to following priority
A1>A2>A3>A4 (10)
Action executing mode is as follows in corresponding Supervised Control pattern:
1. unmanned plane need not interact, when the action under management by exception pattern L3, then allowing to perform is autonomous obstacle
Avoiding action A1, wait instruction action of slowing down A2, spiraling waits instruction action A3 and return action A4, and unmanned plane is according to formula
(10) priority uses autonomous executive mode;
2. unmanned plane need not interact, when in the case where agreeing to management mode L2, it is allowed to which the action of execution refers to for deceleration wait
Order action A2, spiraling waits instruction action A3 and return action A4, and unmanned plane is according to formula (10) priority using the autonomous side of execution
Formula;
3. unmanned plane need not interact, when in the case where instructing control model L1, it is allowed to which the action of execution is return action A4;
Unmanned plane uses autonomous executive mode according to formula (10) priority;
Step4:The environmental factor of Step1 of renewal and the value of obstacle avoidance action, unmanned plane Supervised Control mould per 30s
If formula reduces, renewal unmanned plane Supervised Control pattern is the pattern after reduction, is moved in circles according to Step1~Step3, nobody
Machine performs the obstacle avoidance finally obtained and acts, completion obstacle avoidance, after success obstacle avoidance, unmanned plane Supervised Control pattern
Revert to management by exception pattern L3.
Case of the unmanned plane in obstacle avoidance can be normal by communication and communication disruption is divided into two major classes, meanwhile, communication
It can be divided into again in face of two kinds of situations of known obstacle and unknown obstacle, separately below to the process of above-mentioned three kinds of situations when normal
Emulated, as shown in table 5:
5 unmanned plane obstacle avoidance of table
Embodiment shown in table 5 is divided into the normal and two major class situation of communication disruption that communicates, and communicates and divide under normal circumstances
For known obstacle and unknown obstacle two types.The simulation process of tasks carrying to unmanned plane under three kinds of embodiments below
It is described:
Embodiment 1
(1)Step1:Environmental factor and the expression of obstacle avoidance action
As shown in Table 5, the barrer types are known obstacle, then V1=1, V4=0;Distance of obstacle is that distance is remote, then V3=0,
V5=1;Communication conditions are normal for communication, then V2=0, and unmanned plane action A1~A4 is initialized as 0, in addition, initial supervision control
Molding formula is management by exception pattern L3;
(2)Step2:The reasoning and judging of optional obstacle avoidance action
State vector is:
By its iteration, the state vector of output node is:
Output the result shows that unmanned plane selectable action is autonomous obstacle avoidance A1 in embodiment 1, therefore for known
Obstacle, unmanned plane can give full play to its capacity of will, complete independently obstacle avoidance.
(3)Step3:Final obstacle avoidance action judges
For embodiment 1, the barrer types are known obstacle, and unmanned plane has the ability of autonomous obstacle avoidance, and is making an exception
Under management mode L3, unmanned plane has the authority for performing autonomous obstacle avoidance action, so need not be with there is man-machine handed over
Mutually, unmanned plane can independently complete obstacle avoidance process, and process is as shown in Figure 3.
Case 2
(1)Step1:Environmental factor and the expression of obstacle avoidance action
As shown in Table 5, the barrer types are unknown obstacle, then V1=0, V4=1;Distance of obstacle is that distance is near, then V3=1,
V5=0;Communication conditions are normal for communication, then V2=0, and unmanned plane action A1~A4 is initialized as 0, initial Supervised Control mould
Formula is management by exception pattern.
(2)Step2:The reasoning and judging of optional obstacle avoidance action
State vector is:
By its iteration, the state vector of output node is:
Output the result shows that unmanned plane in example 2, unmanned plane can not automatic obstacle avoiding, then perform spiral wait instruction
Act A3.
(3)Step3:Final obstacle avoidance action judges
Unmanned plane independently cannot complete obstacle avoidance, it is necessary to have it is man-machine interact, while pass back obstacle image letter
Breath and positional information, there is information of the man-machine operator according to barrier, regeneration barrier data, i.e., by unknown complaint message more
It is newly known complaint message.
The environmental factor of Step1 and the expression of obstacle avoidance action are updated, the barrer types are known obstacle, then V1=1,
V4=0;Distance of obstacle is that distance is near, then V3=1, V5=0;Communication conditions are normal for communication, then V2=0, unmanned plane action A1
~A4 is initialized as 0, and Supervised Control pattern is management by exception pattern L3.
Update the reasoning and judging of the optional obstacle avoidance action of Step2
State vector is:
By its iteration, the state vector of output is:
Output is the result shows that the action that unmanned plane selects in the case of scenario 2 is autonomous obstacle avoidance A1.
Update Step3:Final obstacle avoidance action judges
For embodiment 2, the barrer types are unknown obstacle, and unmanned plane does not have the ability of autonomous obstacle avoidance, so needing
With there is man-machine interact;There are the obstacle image information and positional information of man-machine reception unmanned plane passback, complaint message is carried out
Renewal, is allowed to type and is changed into known obstacle;Unmanned plane carries out optional obstacle avoidance further according to the ambient condition under the present situation and moves
The reasoning and judging of work, the barrer types are known obstacle at this time, and unmanned plane has the ability of autonomous obstacle avoidance, and is managed in exception
Under reason pattern L3, unmanned plane has the authority for performing the action of autonomous obstacle avoidance, thus need not with have it is man-machine interact,
Unmanned plane can independently complete obstacle avoidance process, and process is as shown in Figure 3.
Embodiment 3
(1)Step1:Environmental factor and the expression of obstacle avoidance action
By embodiment 3, the barrer types are unknown obstacle, then V1=0, V4=1;Distance of obstacle is that distance is remote, then V3=0,
V5=1;Communication conditions are communication disruption, then V2=1, and unmanned plane action A1~A4 is initialized as 0, and Supervised Control pattern is
Management by exception pattern L3.
(2)Step2:The reasoning and judging of optional obstacle avoidance action
State vector is:
By its iteration, the state vector of output node is:
Output the result shows that unmanned plane in the case of example 3, unmanned plane does not have the ability of automatic obstacle avoiding, slows down then
Instruction A2 is waited, and due to communication disruption, man-machine command can not have been received, to ensure inherently safe, performed and return to operation
A4。
(3)Step3:Final obstacle avoidance action judges
Unmanned plane cannot independently complete obstacle avoidance decision-making, simultaneously because communication disruption, and can not with there is human-computer interaction,
Unmanned plane selection at this time adjusts Supervised Control pattern, Supervised Control pattern is reduced after 15s to operator decision-making mode L2, hereafter
15s still communication disruptions, can not interact, and unmanned plane reduces Supervised Control pattern to control model L1 is instructed, and hereafter 15s is still
Communication disruption, can not interact, and unmanned plane selects self-insurance behavior, come back to the base, that is, perform return according to instruction control model L1
Base acts A4, and process is as shown in Figure 4.
Claims (1)
1. a kind of have the man-machine method for supervision and control to unmanned plane obstacle avoidance, it is characterised in that comprises the following steps:
Step1:Define environmental factor, obstacle avoidance action and Supervised Control pattern
Unmanned plane has 4 kinds of obstacle avoidance actions, including autonomous obstacle avoidance action A1, wait instruction of slowing down when running into barrier
Action A2, spiraling waits instruction action A3 and return action A4;
Wherein, it is high-grade action to define autonomous obstacle avoidance action A1;Slow down to wait instruction A2 and spiral and wait instruction A3
For the action of middle grade;Return to the action that A4 is inferior grade;
Supervised Control pattern is 3 patterns, and the capacity of will of unmanned plane is respectively management by exception pattern L3 from high to low, agrees to pipe
Reason pattern L2 and instruction control model L1;Outer management mode L3 exemplified by the initial pattern of unmanned plane;
It is as shown in table 1 to define unmanned plane Supervised Control pattern;
1 three kinds of Supervised Control patterns of table and its respective action
Action executing mode refers to unmanned plane in the case where there is man-machine Supervised Control, completes the mode of corresponding obstacle avoidance action, such as
Shown in table 2:
2 three kinds of Supervised Control patterns of table and its corresponding action executing mode
1) environmental factor includes that obstacle distance is far and near, whether obstacle information is known and current communication status whether good 3
A aspect, is defined as follows:
1. the definition of obstacle distance, including the nearly V3 of the distance of obstacle and remote V5 of distance of obstacle;
Relative to unmanned plane, the position of barrier can be expressed as
In formula (1), P (t) is the position of t moment barrier i, and t is the time, and P (0) is the initial position of barrier i, viFor obstacle
The speed of thing i, vu(τ) is the speed of unmanned plane u;
In formula (2)For the yaw angle of unmanned plane, unmanned plane and obstacle distance it is far and near by arrival Obstacle Position when
Between T defined by formula below:
2. whether barrier is known to be divided into known obstacle V1 and unknown obstacle V4, it is known that obstacle refers to the barrier in operating area
Image information and positional information are it is known that and unknown obstacle is that the image information of barrier and positional information be extremely in operating area
A rare information is unknown;
3. current communication status is divided into communication normally and interrupts V2, the signal strength that communication disruption is pointed to up to unmanned plane receiver is small
In the sensitivity for having man-machine receiver, cause to have it is man-machine between unmanned plane can not normal communication situation;
2) codomain of environmental factor and the value of obstacle avoidance action is { 0,1 }, wherein 1 represents that the state is effective, 0 represents state
It is invalid;Specific environment factor and the value of operating state such as table 3 below, shown in table 4:
The value of 3 environmental factor of table
Wherein * represents outlier, and outlier represents that value can be " 0 " or " 1 ", and the value does not influence corresponding state under *, only
Influencing it needs to take the state under particular value;
The value of 4 unmanned plane operating state of table
Step2:The reasoning and judging of the optional obstacle avoidance action of unmanned plane
1) state vector is write out by Step1
State vector represents that wherein k represents number with C (k), and C (0) represents initial state vector, and state vector includes obstacle
Avoiding action and the full content of environmental factor, form are:
Wherein, during no avoiding action, the value of acquiescence unmanned plane A1~A4 is all 0, i.e.,
V1~V5 is the value that Step1 is obtained, and is always definite value in Step2;
2) made inferences using state vector C (k), reasoning process is as follows:
1. multiplying adjacent weight matrix W with state vector C (k) obtains intermediate vector X (k):
X (k)=C (k) W (6)
Wherein, the adjoining weight matrix W between node is as follows:
2. to intermediate vector X (k) each component x of state transition function f (x) processing intermediate vector X (k), state transfer letter
Number f (x) is two-value jump function:
Wherein, x is the component of intermediate vector X (k), and vectorial X (k) dimension is 6, and the codomain of component x is { 0,1 };
3. state vector C (k) is updated with intermediate vector X (k), i.e.,
4. continuous repeat step 1., step 2. with step 3., that is, state vector C (k) is constantly updated, until state vector C (k+
N+1)=C (k+n), wherein, n represents number, the i.e. value of (k+n+1) next state vector and the value of (k+n) next state vector
It is identical;
C (k+n+1)=C (k+n)=[a1, a2, a3, a4 | V1, V2, V3, V4, V5], wherein a1, a2, a3, a4 represent to hinder respectively
Hinder the corresponding value of avoiding action A1, A2, A3, A4, be worth for 0 or 1, value is 1 item in a1, a2, a3, a4, corresponding obstacle
Evade moving and acted for optional obstacle avoidance;
It is man-machine to having that 5. unmanned plane provides optional obstacle avoidance the result of the action;
Step3:Final obstacle avoidance action judges
1) when meeting following condition a) and b) wherein one, then unmanned plane is needed with there is man-machine interact:
A) be " 1 " comprising unknown obstacle V4 entry value in environmental factor, i.e. V=(*, *, *, 1, *), wherein, * represents outlier, nothing
Close item and represent that value is " 0 " or " 1 ";
B) the current Supervised Control pattern respective action grade of unmanned plane is closed in optional obstacle avoidance performance of a different dive;
Action executing mode in current Supervised Control pattern is as follows:
1. when the Supervised Control pattern of unmanned plane is management by exception pattern L3, and unmanned plane needs interactive, as in 15s unmanned plane without
The obstacle avoidance action reasoning that method provides is as a result, i.e. barrier is unknown, then unmanned plane reduces Supervised Control pattern and managed to agreement
Pattern L2;If unmanned plane is to having man-machine offer obstacle avoidance action reasoning as a result, unmanned plane is actively anti-using intelligent combination
Feedback, operator do not negate the obstacle avoidance action reasoning for performing unmanned plane offer in 15s as a result, conversely, operator is in 15s
Interior negative does not perform then, and reduces Supervised Control pattern to agreement management mode L2;
2. when the Supervised Control pattern of unmanned plane is agrees to management mode L2, when unmanned plane needs interactive, as in 15s unmanned plane without
The obstacle avoidance action reasoning that method provides is as a result, i.e. barrier is unknown, then unmanned plane reduces Supervised Control pattern and managed to agreement
Pattern L1;If unmanned plane is to having man-machine offer obstacle avoidance action reasoning as a result, unmanned plane is actively anti-using intelligent combination
Feedback, waits the accreditation of operator, and operator does not negate then to perform the obstacle avoidance action reasoning knot that unmanned plane provides in 15s
Fruit, the interior no negatives of 15s do not perform then, and reduce Supervised Control pattern to operator's decision-making mode L1;
3. when the Supervised Control pattern of unmanned plane is operator decision-making mode L1, and unmanned plane needs interactive, unmanned plane is to someone
Machine provides the image information and positional information of barrier, and state vector C (k) and obstacle avoidance action reasoning are provided as a result,
There is man-machine combined using pipe-connecting mode to perform:
If (known to the barrer types), then (select autonomous obstacle avoidance decision-making to evade)
If (the barrer types are unknown and apart from remote), then (selection, which is slowed down, waits instruction)
If (the barrer types are unknown and apart near), then (selection, which is spiraled, waits instruction) (9)
Unmanned plane obstacle avoidance is provided by order to act, and the decision-making of operator is waited in unmanned plane 15s, is held according to the result of decision
OK, time-out then comes back to the base without decision-making;
2) the current Supervised Control pattern respective action grade of unmanned plane is met when optional obstacle avoidance acts, and environmental factor is not
When there is the situation of the Rule of judgment a) of step 1) in step2, unmanned plane is not required to and has human-computer interaction, optional obstacle avoidance action
Performed according to following priority:
A1 > A2 > A3 > A4 (10)
Action executing mode is as follows in corresponding Supervised Control pattern:
1. unmanned plane need not interact, when the action under management by exception pattern L3, then allowing to perform is autonomous obstacle avoidance
Action A1, wait instruction action of slowing down A2, spiraling waits instruction action A3 and return action A4, and unmanned plane is preferential according to formula (10)
Level uses autonomous executive mode;
2. unmanned plane need not interact, when in the case where agreeing to management mode L2, it is allowed to which the action of execution waits instruction dynamic to slow down
Make A2, spiraling waits instruction action A3 and return action A4, unmanned plane uses autonomous executive mode according to formula (10) priority;
3. unmanned plane need not interact, when in the case where instructing control model L1, it is allowed to which the action of execution is return action A4;Nobody
Machine uses autonomous executive mode according to formula (10) priority;
Step4:The environmental factor of Step1 of renewal and the value of obstacle avoidance action per 30s, if unmanned plane Supervised Control pattern
Reduce, renewal unmanned plane Supervised Control pattern is the pattern after reduction, is moved in circles according to Step1~Step3, and unmanned plane performs
The obstacle avoidance finally obtained acts, completion obstacle avoidance, after success obstacle avoidance, exemplified by unmanned plane Supervised Control pattern recovery
Outer management mode L3.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710553624.XA CN107943066B (en) | 2017-07-08 | 2017-07-08 | Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710553624.XA CN107943066B (en) | 2017-07-08 | 2017-07-08 | Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107943066A true CN107943066A (en) | 2018-04-20 |
CN107943066B CN107943066B (en) | 2020-01-14 |
Family
ID=61928567
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710553624.XA Active CN107943066B (en) | 2017-07-08 | 2017-07-08 | Method for supervising and controlling obstacle avoidance of unmanned aerial vehicle by using human |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107943066B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020000790A1 (en) * | 2018-06-29 | 2020-01-02 | 太原理工大学 | Vertical mine shaft detection method and system |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5581250A (en) * | 1995-02-24 | 1996-12-03 | Khvilivitzky; Alexander | Visual collision avoidance system for unmanned aerial vehicles |
EP2674723A2 (en) * | 2012-06-11 | 2013-12-18 | Honeywell International Inc. | Systems and methods for unmanned aircraft system collision avoidance |
CN103592948A (en) * | 2013-12-04 | 2014-02-19 | 成都纵横自动化技术有限公司 | Unmanned aerial vehicle flying anti-collision method |
CN103699132A (en) * | 2013-12-05 | 2014-04-02 | 中国航空无线电电子研究所 | Device and method for assisting visual disc to precess and approach |
CN106094569A (en) * | 2016-07-06 | 2016-11-09 | 西北工业大学 | Multi-sensor Fusion unmanned plane perception with evade analogue system and emulation mode thereof |
US20170102241A1 (en) * | 2014-03-15 | 2017-04-13 | Aurora Flight Sciences Corporation | Autonomous Vehicle Navigation System and Method |
CN106647810A (en) * | 2017-01-10 | 2017-05-10 | 山东科技大学 | UAV automatic collision avoidance method based on negative-proportion guiding |
-
2017
- 2017-07-08 CN CN201710553624.XA patent/CN107943066B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5581250A (en) * | 1995-02-24 | 1996-12-03 | Khvilivitzky; Alexander | Visual collision avoidance system for unmanned aerial vehicles |
EP2674723A2 (en) * | 2012-06-11 | 2013-12-18 | Honeywell International Inc. | Systems and methods for unmanned aircraft system collision avoidance |
CN103592948A (en) * | 2013-12-04 | 2014-02-19 | 成都纵横自动化技术有限公司 | Unmanned aerial vehicle flying anti-collision method |
CN103699132A (en) * | 2013-12-05 | 2014-04-02 | 中国航空无线电电子研究所 | Device and method for assisting visual disc to precess and approach |
US20170102241A1 (en) * | 2014-03-15 | 2017-04-13 | Aurora Flight Sciences Corporation | Autonomous Vehicle Navigation System and Method |
CN106094569A (en) * | 2016-07-06 | 2016-11-09 | 西北工业大学 | Multi-sensor Fusion unmanned plane perception with evade analogue system and emulation mode thereof |
CN106647810A (en) * | 2017-01-10 | 2017-05-10 | 山东科技大学 | UAV automatic collision avoidance method based on negative-proportion guiding |
Non-Patent Citations (1)
Title |
---|
魏瑞轩、李学仁: "《先进无人机系统与作战运用》", 30 September 2014 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020000790A1 (en) * | 2018-06-29 | 2020-01-02 | 太原理工大学 | Vertical mine shaft detection method and system |
Also Published As
Publication number | Publication date |
---|---|
CN107943066B (en) | 2020-01-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Santini et al. | Platooning maneuvers in vehicular networks: A distributed and consensus-based approach | |
CN107491085A (en) | It is a kind of to have the man-machine method for supervision and control evaded to unmanned plane threat | |
CN112462602B (en) | Distributed control method for keeping safety spacing of mobile stage fleet under DoS attack | |
Baldi et al. | Platooning merging maneuvers in the presence of parametric uncertainty | |
Huang et al. | Multi-agent vehicle formation control based on mpc and particle swarm optimization algorithm | |
Zhang et al. | Autonomous vehicle with communicative driving for pedestrian crossing: Trajectory optimization | |
CN107943066A (en) | It is a kind of to have the man-machine method for supervision and control to unmanned plane obstacle avoidance | |
Klimke et al. | An enhanced graph representation for machine learning based automatic intersection management | |
WO2020018011A1 (en) | Control of traffic lights that govern vehicular traffic at a junction of roads | |
Tallapragada et al. | Distributed control of vehicle strings under finite-time and safety specifications | |
Chamideh et al. | A safe and robust autonomous intersection management system using a hierarchical control strategy and v2i communication | |
Zennir et al. | Comparison of PID and fuzzy controller for path tracking control of autonomous electrical vehicles | |
Guillen-Perez et al. | Raim: Reinforced autonomous intersection management—aim based on madrl | |
Wang et al. | Multi-smart car control system design and research based on ZigBee | |
Krishnan et al. | Continuous-time trajectory optimization for decentralized multi-robot navigation | |
Zou et al. | Adaptive control of autonomous vehicle platoons | |
Mokhtari et al. | Pedestrian collision avoidance for autonomous vehicles at unsignalized intersection using deep q-network | |
Marinica et al. | Distributed collision avoidance for autonomous vehicles: world automata representation | |
Typaldos et al. | Modified dynamic programming algorithms for GLOSA systems with stochastic signal switching times | |
Wang et al. | Speed planning for autonomous driving in dynamic urban driving scenarios | |
Al-Aubidy et al. | Multi-robot task scheduling and routing using neuro-fuzzy control | |
Honda et al. | Mpc builder for autonomous drive: Automatic generation of mpcs for motion planning and control | |
Gao et al. | Proactive platooning based on C-V2X to relieve congestion at a signalized intersection | |
Németh et al. | Hierarchical control design of automated vehicles for multi-vehicle scenarios in roundabouts | |
EP3916555A1 (en) | Method and device for latency compensation |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20200812 Address after: 100190, No. 9 West Fourth Ring Road, Beijing, Haidian District Patentee after: Beijing Yuanchen Huasheng Technology Co., Ltd Address before: 710072 Xi'an friendship West Road, Shaanxi, No. 127 Patentee before: Northwestern Polytechnical University |