CN112947111A - Machine learning-oriented middle and long distance air action parameter determination method - Google Patents
Machine learning-oriented middle and long distance air action parameter determination method Download PDFInfo
- Publication number
- CN112947111A CN112947111A CN202011599745.6A CN202011599745A CN112947111A CN 112947111 A CN112947111 A CN 112947111A CN 202011599745 A CN202011599745 A CN 202011599745A CN 112947111 A CN112947111 A CN 112947111A
- Authority
- CN
- China
- Prior art keywords
- maneuver
- control
- maneuvering
- machine learning
- parameters
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B17/00—Systems involving the use of models or simulators of said systems
- G05B17/02—Systems involving the use of models or simulators of said systems electric
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The application belongs to the technical field of simulation system design, and particularly relates to a machine learning-oriented middle and long distance air action parameter determination method. Comprising step S1, determining the type of maneuver for the middle-distance air maneuver; s2, simulating by using a maneuvering simulation program to obtain a typical track of each maneuvering action; step S3, according to the operation change situation of the typical track, taking the time corresponding to the operation change exceeding the threshold value as a dividing point, and decomposing the typical track into a plurality of control segments; step S4, obtaining the access condition limiting parameters of each maneuver and the control parameters of each control section; and step S5, adjusting the state of the airplane according to the entry condition during machine learning, and performing flight control calculation according to the control parameters to form a maneuvering control instruction. According to the method and the device, the control strategy and the control parameters are decomposed and the conditions of entering and exiting are judged according to the control logic of the pilot in the battle, so that the relatively real simulation of the support air combat behavior is achieved.
Description
Technical Field
The application belongs to the technical field of simulation system design, and particularly relates to a machine learning-oriented middle and long distance air action parameter determination method.
Background
With the continuous improvement of the performance of the combat aircraft, the situation of the modern air battlefield changes instantly and the combat aircraft needs to flexibly change flight parameters such as flight height, speed, course, gradient and the like in order to take advantage of the situation. In recent years, the intelligent air combat decision problem is more and more emphasized by military operation and weapon development departments of various countries, and the most core content in the air combat decision is the air combat maneuver decision. In the optimization method of the air combat maneuver decision, a maneuver library is firstly designed. The action library comprises action sets for decision selection, and the action library design is the basis of the air combat maneuver decision. In modern air combat simulations, the tactical actions of a pilot in air combat can be viewed as a series of air combat tactical actions arranged in time, i.e., the air combat tactical actions are a chain of pilot tactical actions in the time dimension. In the design of a typical tactical action library, an important problem is a control strategy of a maneuvering action, namely how to realize the fitting of the maneuvering action to a more real air combat behavior, and meanwhile, in intelligent air combat, an intelligent computer more accurately controls and realizes a flight track meeting combat requirements. The current maneuvering actions are various, many maneuvering controls are complex, and the application difference with actual combat fit is large.
The scheme for solving the problems is to combine actual combat experience and operate according to a combat operation control strategy in the pilot combat process to form a relevant air combat maneuver control method.
Disclosure of Invention
The application provides a machine learning-oriented middle and long distance air action parameter determination method, which comprises the following steps:
step S1, determining the maneuvering action type for performing the middle-long-distance air action;
s2, simulating by using a maneuvering simulation program to obtain a typical track of each maneuvering action;
step S3, according to the operation change situation of the typical track, taking the time corresponding to the operation change exceeding the threshold value as a dividing point, and decomposing the typical track into a plurality of control segments;
step S4, obtaining the access condition limiting parameters of each maneuver and the control parameters of each control section;
and step S5, adjusting the state of the airplane according to the entering conditions during machine learning, and performing flight control calculation according to the control parameters to form a maneuvering control instruction.
Preferably, after step S1, the method further includes obtaining the maneuver description and tactical intention of each maneuver, and solidifying into a database file.
Preferably, in step S3, the exemplary trajectory is decomposed into at least an entry segment, an intermediate segment, and a departure segment.
Preferably, the control parameters include a track-determined heading, a track plane inclination, a throttle lever slope, a longitudinal acceleration, and an overload value.
Preferably, the throttle lever slope is determined in accordance with the trajectory plane inclination.
Preferably, the maneuvering action types include constant speed leveling, offset, leveling flight acceleration and deceleration, downward weighing bucket, fast climbing, S-shaped, inclined weighing bucket/half inclined weighing bucket, tracking guidance, horizontal turning, fast turning, stable turning, elevation shooting, diving maneuvering, offset diving, diving maneuvering and reversing, and half rolling reversing.
The method and the device take the purposes of operation use and tactics as starting points, decompose control strategies and control parameters and judge the entering and exiting conditions according to the control logic of the pilot in the operation so as to achieve the more real simulation of the support air combat behavior.
Drawings
Fig. 1 is a flowchart of a preferred embodiment of the method for determining air motion parameters at a middle and long distance for machine learning.
Fig. 2 is a schematic diagram of a typical trajectory of a bucket jack-down maneuver.
Fig. 3 is a schematic diagram of the variation of the height of a typical parameter of a bucket-down maneuver.
Detailed Description
In order to make the implementation objects, technical solutions and advantages of the present application clearer, the technical solutions in the embodiments of the present application will be described in more detail below with reference to the accompanying drawings in the embodiments of the present application. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are some, but not all embodiments of the present application. The embodiments described below with reference to the drawings are illustrative and intended to be used for explaining the present application and should not be construed as limiting the present application. All other embodiments that can be derived by a person skilled in the art from the embodiments given herein without making any creative effort fall within the protection scope of the present application. Embodiments of the present application will be described in detail below with reference to the drawings.
The application provides a machine learning-oriented method for determining air motion parameters at a middle and long distance, as shown in fig. 1, which mainly comprises the following steps:
step S1, determining the maneuvering action type for performing the middle-long-distance air action;
s2, simulating by using a maneuvering simulation program to obtain a typical track of each maneuvering action;
step S3, according to the operation change situation of the typical track, taking the time corresponding to the operation change exceeding the threshold value as a dividing point, and decomposing the typical track into a plurality of control segments;
step S4, obtaining the access condition limiting parameters of each maneuver and the control parameters of each control section;
and step S5, adjusting the state of the airplane according to the entering conditions during machine learning, and performing flight control calculation according to the control parameters to form a maneuvering control instruction.
In some alternative embodiments, after step S1, the method further includes obtaining a maneuver description and a tactical intention of each maneuver, and solidifying into a database file.
In some alternative embodiments, in step S3, the exemplary trajectory is decomposed into at least an entry segment, an intermediate segment, and a departure segment.
In some alternative embodiments, the control parameters include a track-determined heading, a track plane tilt, a throttle lever grade, a longitudinal acceleration, and an overload value.
The details are as follows.
Determination of maneuver
According to the public air combat description knowledge and the common flight maneuver of pilots, 16 typical maneuvers including constant speed flat flight, flat flight acceleration and deceleration, fast climbing, tilting bucket/semi-tilting bucket, horizontal turning, stable turning, diving maneuver transformation, offset, downward-lifting bucket, S-shaped, tracking guidance, fast turning, elevation shooting, offset diving and semi-rolling reversing are summarized and summarized. Typical motor actions are shown in table 1.
TABLE 1 typical maneuver
| Serial number | Name (R) | Serial number | Name (R) |
| 1 | Constant speed flat fly | 9 | Biasing |
| 2 | Plane flight acceleration and deceleration | 10 | Bucket for lifting weight downwards |
| 3 | Quick climbing device | 11 | S shape |
| 4 | Tilted-catty bucket/semi-tilted-catty bucket | 12 | Tracking guidance |
| 5 | Horizontal turn | 13 | Quick turn |
| 6 | Stable turning | 14 | Upward-shooting |
| 7 | Diving maneuver | 15 | Offset nose-down |
| 8 | Motor-driven extension of diving | 16 | Semi-rolling reverse rotation |
(II) description of the Process
The description of the maneuver includes the following:
description of the maneuver: the name of the maneuver, which gives more detailed definition and description of the maneuver;
tactical intentions: describing the fighting intention, fighting scene and tactical purpose of the mobile action;
typical trajectories: simulating and analyzing a typical track of the maneuvering action by using a maneuvering simulation program;
the operation process is decomposed: aiming at the realization of the maneuvering action, the operation steps and the operation key of the accelerator and the steering column are described.
Controlling parameters: and (4) giving control parameter suggestions through simulation analysis.
(III) description flow of maneuver is shown in FIG. 1:
when designing the over-the-horizon air combat action library, the actions are taken first according to the requirements. The selected actions are familiar with their basic form, as well as their tactical function. The tactical action is specifically designed, namely, the value sequence of the control quantity is set according to the rule determined by the action form and the tactical function thereof. And determining required control quantity values and ending conditions required by each action in the action sequence. I.e. determining the gradient, longitudinal acceleration and overload values required to perform each action in the sequence of actions, called motion demand values. Each action has a specific requirement on the control quantity, and the required value of the control quantity can be determined by establishing an action library. The conditions for executing the action end include time, speed, height difference, angle and the like.
The method takes the operational use and tactical purposes as the starting point, forms a maneuvering action library which meets the actual combat requirements and is more simplified and efficient in control, decomposes control strategies and control parameters according to the control logic of pilots in actual combat, judges the entering and exiting conditions, forms a description that the aerodynamic behavior of the air combat is comprehensive and close to the actual combat, is more beneficial to being applied in an intelligent air combat system, and is more scientific and reasonable.
For example.
The maneuvering name is as follows: bucket for lifting weight downwards
Description of (II) Motor vehicle
And (3) finishing the downward bucket in the appointed inclined plane, and describing the implementation method:
a) downward turning is implemented, the vehicle descends for a certain height while the change of the course direction is finished by 180 degrees, and after the avoidance is finished, the tail placement is finished upwards;
b) the maneuvering track is in an inclined plane, and the track in the inclined plane is in a circular arc shape.
Tactical intentions (III)
The front half section of the downward bucket is used for avoiding missiles launched by enemy aircraft, after the successful avoidance is confirmed, the rear half section is executed, the missile is revolved and climbed to occupy favorable situations, and the missiles are launched. A jack-down bucket is a maneuverable action that provides both defense and attack.
(IV) typical trajectories
The typical trajectory of the downward bucket maneuvering is shown in fig. 2, the typical parameter change of the downward bucket maneuvering is shown in fig. 3, fig. 3 only shows the change situation of the height, and in the actual operation process, the change situations of the speed, the overload, the attack angle, the pitch angle, the roll angle and the mach number should be given.
(V) decomposition of operation process
Entry condition determination
Before entering the downward weighing hopper, whether the current flight state parameters meet the maneuvering condition or not and whether state adjustment is needed or not are judged.
Process segment 1: selecting a turning direction (left/right), adjusting the accelerator to adjust the state of the engine to a given state, establishing a required gradient to a turning direction pressure lever, placing a driving lever at a neutral position without sideslip, and stably establishing overload of 4-5g within 2-3 s;
a process section 2: the tension rod keeps the overload of 4-5g (when the attack angle reaches the limit of the attack angle, the attack angle is kept), and the amplitude of the compression rod is adjusted according to the course change.
Process section 3: the tension rod keeps the overload of 4-5g (when the attack angle reaches the attack angle limit, the attack angle is kept), the amplitude of the compression rod is adjusted according to the course change, and the course is controlled to the initial course.
A modification section: the rod returns to adjust the gradient to level and keep level flight.
(VI) control the process decomposition, see Table 2 below.
TABLE 2 control Process decomposition
(VII) control parameter selection
The control parameters of the bucket weight are shown in the table 3:
TABLE 3 bucket control parameters
And finally, during machine learning, adjusting the state of the airplane according to the entry condition, and performing flight control calculation according to the control parameters to form a maneuvering control instruction.
Having thus described the present invention with reference to the preferred embodiments illustrated in the accompanying drawings, it will be understood by those skilled in the art that the scope of the present invention is not limited to those specific embodiments, and that equivalent modifications or substitutions of the related technical features may be made by those skilled in the art without departing from the principle of the present invention, and those modifications or substitutions will fall within the scope of the present invention.
Claims (6)
1. A method for determining air motion parameters at a middle and long distance facing machine learning is characterized by comprising the following steps:
step S1, determining the maneuvering action type for performing the middle-long-distance air action;
s2, simulating by using a maneuvering simulation program to obtain a typical track of each maneuvering action;
step S3, according to the operation change situation of the typical track, taking the time corresponding to the operation change exceeding the threshold value as a dividing point, and decomposing the typical track into a plurality of control segments;
step S4, obtaining the access condition limiting parameters of each maneuver and the control parameters of each control section;
and step S5, adjusting the state of the airplane according to the entry condition during machine learning, and performing flight control calculation according to the control parameters to form a maneuvering control instruction.
2. The method for determining parameters of long-distance air activities oriented to machine learning of claim 1, wherein after step S1, further comprising obtaining maneuver description and tactical intention of each maneuver and solidifying into database file.
3. The machine learning-oriented medium-distance air motion parameter determination method according to claim 1, wherein in step S3, the typical trajectory decomposition is at least decomposed into an entry segment, a middle segment and a departure segment.
4. The machine-learning oriented medium-distance air motion parameter determination method of claim 1, wherein the control parameters include a track-determined heading, a track plane tilt angle, a throttle lever slope, a longitudinal acceleration, and an overload value.
5. The machine-learning-oriented medium-distance air motion parameter determination method according to claim 4, wherein the throttle lever gradient is determined in accordance with the trajectory plane inclination angle.
6. The machine learning oriented medium-distance airborne maneuver parameter determination method of claim 1, wherein the maneuver types include equal speed level, offset, level flight acceleration and deceleration, jack down, fast climb, S-shape, jack-up/jack-up, tracking guidance, level turn, fast turn, steady turn, pitch maneuver, offset pitch, pitch maneuver pull-out, and half-roll reverse.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011599745.6A CN112947111A (en) | 2020-12-29 | 2020-12-29 | Machine learning-oriented middle and long distance air action parameter determination method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011599745.6A CN112947111A (en) | 2020-12-29 | 2020-12-29 | Machine learning-oriented middle and long distance air action parameter determination method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112947111A true CN112947111A (en) | 2021-06-11 |
Family
ID=76234966
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011599745.6A Pending CN112947111A (en) | 2020-12-29 | 2020-12-29 | Machine learning-oriented middle and long distance air action parameter determination method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112947111A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114239281A (en) * | 2021-12-17 | 2022-03-25 | 中国航空研究院 | Battlefield information ontology model construction method for multi-domain cooperative combat |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108319286A (en) * | 2018-03-12 | 2018-07-24 | 西北工业大学 | A kind of unmanned plane Air Combat Maneuvering Decision Method based on intensified learning |
| CN112130457A (en) * | 2020-09-21 | 2020-12-25 | 南京航空航天大学 | Fuzzy flight control method for variant unmanned aerial vehicle perching and landing maneuver |
-
2020
- 2020-12-29 CN CN202011599745.6A patent/CN112947111A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108319286A (en) * | 2018-03-12 | 2018-07-24 | 西北工业大学 | A kind of unmanned plane Air Combat Maneuvering Decision Method based on intensified learning |
| CN112130457A (en) * | 2020-09-21 | 2020-12-25 | 南京航空航天大学 | Fuzzy flight control method for variant unmanned aerial vehicle perching and landing maneuver |
Non-Patent Citations (1)
| Title |
|---|
| 刘瑛: "复杂机动动作最优航迹控制模型及操纵特性分析", 《控制理论与应用》, pages 566 - 576 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114239281A (en) * | 2021-12-17 | 2022-03-25 | 中国航空研究院 | Battlefield information ontology model construction method for multi-domain cooperative combat |
| CN114239281B (en) * | 2021-12-17 | 2024-05-03 | 中国航空研究院 | Battlefield information ontology model construction method for multi-domain collaborative combat |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108168381B (en) | A kind of control method of more pieces of guided missile cooperations | |
| Park et al. | Differential game based air combat maneuver generation using scoring function matrix | |
| CN101286071B (en) | Multiple no-manned plane three-dimensional formation reconfiguration method based on particle swarm optimization and genetic algorithm | |
| Austin et al. | Automated maneuvering decisions for air-to-air combat | |
| CN107063255B (en) | Three-dimensional route planning method based on improved drosophila optimization algorithm | |
| CN108319286A (en) | A kind of unmanned plane Air Combat Maneuvering Decision Method based on intensified learning | |
| CN111473696B (en) | Guidance rocket vertical attack guidance method based on drop point estimation | |
| CN105204512B (en) | A kind of six degree of freedom UCAV short range combat method based on simplified model game playing by machine | |
| US9711052B2 (en) | Flight path search device and flight path search program | |
| CN111666631A (en) | Unmanned aerial vehicle maneuvering decision method combining hesitation fuzzy and dynamic deep reinforcement learning | |
| CN110187640B (en) | Multi-missile cooperative combat guidance law design method for maneuvering target and allowable communication time lag | |
| CN114200826B (en) | High stability control method for supersonic speed large maneuvering target continuous large overload maneuvering | |
| CN110822994B (en) | Linear pseudo-spectrum spreading control guidance method with falling angle constraint | |
| CN109737812B (en) | Air-to-ground guided weapon side attack method and device | |
| CN114492805A (en) | Air combat maneuver decision design method based on fuzzy reasoning | |
| CN113282061A (en) | Unmanned aerial vehicle air game countermeasure solving method based on course learning | |
| CN113625740A (en) | Unmanned aerial vehicle air combat game method based on transfer learning pigeon swarm optimization | |
| CN114115332A (en) | Ground sweeping flight technology applied to short-range supersonic speed cruise bomb | |
| CN112947111A (en) | Machine learning-oriented middle and long distance air action parameter determination method | |
| CN115903865A (en) | Aircraft near-distance air combat maneuver decision implementation method | |
| CN114756959A (en) | Design method of aircraft short-distance air combat maneuver intelligent decision machine model | |
| CN114637319A (en) | Semi-rolling reverse maneuvering flight control method of two-channel unmanned aerial vehicle | |
| CN113741186B (en) | Double-aircraft air combat decision-making method based on near-end strategy optimization | |
| CN111176315A (en) | Variant patrol missile control method based on L1 adaptive control | |
| Kang et al. | Beyond-visual-range tactical game strategy for multiple uavs |
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 |