CN-121995326-A - Self-adaptive decision method based on multi-unmanned aerial vehicle cooperative interference

CN121995326ACN 121995326 ACN121995326 ACN 121995326ACN-121995326-A

Abstract

The invention provides a self-adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference, which comprises the steps of constructing a multi-unmanned aerial vehicle cooperative interference task decision-making model under an electronic countermeasure scene comprising a radar and a plurality of unmanned aerial vehicles, carrying out parameterization modeling on interference behaviors of the interfering unmanned aerial vehicles based on the multi-unmanned aerial vehicle cooperative interference task decision-making model to generate an interference action space, training the multi-unmanned aerial vehicle cooperative interference task decision-making model based on the interference action space to obtain a cooperative interference decision-making model, deploying the cooperative interference decision-making model on each interfering unmanned aerial vehicle, extracting radar signal characteristics and constructing a local observation state by each interfering unmanned aerial vehicle only based on radar radiation signals intercepted by an own-mounted electronic support measure system, and independently outputting interference decisions under the condition of no centralized control and no explicit communication so as to complete a cooperative interference task.

Inventors

LUO JIA
XIAO HENG
XU CHUAN
LUO HAO
WANG YANG
Yao Haonan

Assignees

重庆邮电大学

Dates

Publication Date: 20260508
Application Date: 20260116

Claims (9)

1. A self-adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference is characterized by comprising the steps of constructing a multi-unmanned aerial vehicle cooperative interference task decision-making model under an electronic countermeasure scene comprising a radar and a plurality of unmanned aerial vehicles, carrying out parameterization modeling on interference behaviors of the interfering unmanned aerial vehicles based on the multi-unmanned aerial vehicle cooperative interference task decision-making model to generate an interference action space, training the multi-unmanned aerial vehicle cooperative interference task decision-making model based on the interference action space to obtain a cooperative interference decision-making model, deploying the cooperative interference decision-making model to each interfering unmanned aerial vehicle, extracting radar signal characteristics and constructing a local observation state by each interfering unmanned aerial vehicle only based on radar radiation signals intercepted by an own-vehicle electronic support measure system, and independently outputting interference decisions under the condition of no centralized control and no explicit communication, so that a cooperative interference task is completed.
2. The self-adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference is characterized by comprising the steps of modeling each interfering unmanned aerial vehicle as an independent decision-making agent, modeling the multi-unmanned aerial vehicle cooperative interference task as a part of observable Markov game model, setting a radar working mode, wherein the radar working mode comprises a search mode, a tracking mode and a confirmation mode, passively intercepting radar radiation signals through an airborne electronic support measure system according to the working mode, obtaining an arrival angle, working frequency, pulse width, pulse repetition interval and received signal power of the radar signals, and constructing a local observation state of the interfering unmanned aerial vehicle.
3. The adaptive decision method based on multi-unmanned aerial vehicle cooperative interference according to claim 2, wherein the search mode is as follows: Regular swing, power exhibiting sweep fluctuation, confirmation mode: short stay at local, fluctuation weakening, tracking mode: basically stable, and the power is continuously stable and high.
4. The adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference according to claim 1, wherein the parameterized modeling of the interference behavior of the interfering unmanned aerial vehicle comprises the steps of representing the interference behavior of the interfering unmanned aerial vehicle at any decision time as an interference action consisting of discrete actions and continuous parameters, and calculating the actual transmitting power of the interfering unmanned aerial vehicle according to the interference action.
5. The adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference of claim 4, wherein the actual transmitting power of the interfering unmanned aerial vehicle is calculated as follows: ; Wherein, the And Respectively representing the maximum and minimum transmitting power of the jammer; representing successive values of the jammer power level, normalized interference power level.
6. The adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference according to claim 1, wherein the centralized training of the multi-unmanned aerial vehicle cooperative interference task decision-making model comprises the steps of obtaining a training set, introducing global observable information of the environment in a training stage, and constructing a radar real working state Wherein 0 represents a search state, 1 represents a confirmation state, and 2 represents a tracking state, acquiring a radar position Maintaining a relatively stable, radar threat level in a confirmed or tracked state Information and location of each interfering drone The method comprises the steps of establishing a system-level reward function based on a global state of information and a multi-unmanned aerial vehicle combined interference action, wherein the uncertainty is increased to reflect effective interference of an interference radar decision process when a radar working mode is frequently switched, a beam direction is unstable or a behavior characteristic is difficult to predict due to interference, and the reward function is used for guiding a learning process of a multi-unmanned aerial vehicle collaborative interference strategy.
7. The adaptive decision method based on multi-unmanned aerial vehicle cooperative interference of claim 6, wherein the system-level reward function is: ; ; ; ; ; Wherein, the In order to represent the radar locking penalty, A penalty for the consumption of resources is given, To reward the probability of success for the disturbance, For probability distribution estimation of radar behavior under historical observation, Entropy is not determined for the radar's behaviour at time t, As a function of the interference utility, For the uncertainty the weighting coefficients are awarded, As a variation of the uncertainty of the radar behaviour, 、 As the weight coefficient of the light-emitting diode, For radar to receive signal under interference the change in the noise ratio relative to the amount of interference free, Representing the residence time of the radar in a particular direction or the amount of change in tracking behavior relative to the historical mean.
8. The adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference of claim 6, wherein the value evaluation of the multi-unmanned aerial vehicle joint interference behavior comprises constructing a joint cost function based on a global state and multi-unmanned aerial vehicle joint interference actions, wherein the network uses the local actions of all the intelligent agents Value and global state The method comprises the steps of inputting and outputting global action value, constraining the joint action value function to meet monotonicity bars for individual action value functions of each interference unmanned aerial vehicle, constructing an interference utility function based on radar signal-to-noise ratio change, radar tracking performance change and radar behavior uncertainty change, quantitatively evaluating joint interference effect of multiple unmanned aerial vehicles, and when marginal gain of interference utility function to interference power is lower than a preset threshold value And when the corresponding interference behavior is judged to be redundant interference, and the weight of the corresponding interference behavior in the joint action value evaluation is reduced.
9. The adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference according to claim 1, wherein the adaptive decision-making method based on multi-unmanned aerial vehicle cooperative interference is characterized in that in the distributed execution stage, each interfering unmanned aerial vehicle only builds a local observation state based on radar radiation signals passively intercepted by an onboard electronic support measure system of the interfering unmanned aerial vehicle, and the unmanned aerial vehicle is based on a cooperative interference decision-making model obtained in the centralized training stage and according to the local observation state The method comprises the steps of independently calculating interference actions, carrying out explicit communication of state information, action information or value information among the interference unmanned aerial vehicles, combining the interference actions of the interference unmanned aerial vehicles to form combined interference actions of a system, carrying out centralized training and parameter updating, and obtaining strategy consistency of implicit learning of a cooperative interference effect in a centralized training stage.

Description

Self-adaptive decision method based on multi-unmanned aerial vehicle cooperative interference Technical Field The invention belongs to the technical field of unmanned aerial vehicle cluster intelligent decision-electromagnetic spectrum countermeasure-cooperative control intersection, and particularly relates to a self-adaptive decision method based on multi-unmanned aerial vehicle cooperative interference. Background With the development of technology, the battlefield of military countermeasure has been expanded from the three-dimensional outside of the sea, land, air and air, and the electromagnetic space becomes the hot spot field of the current large national countermeasure. The increasing intellectualization and informatization of military weapons means that the win or lose of the electronic countermeasure field directly affects the war trend, and whether the advantage can be obtained in the electronic countermeasure has a crucial effect on the final win or lose. Electronic countermeasure means that electromagnetic energy is used, so that enemy electronic equipment is difficult to work normally while protecting the equipment, and the electronic countermeasure is a main form of current information war. The radar side and the interference side are used as the front side and the back side in the countermeasure, and the technologies of the two sides are continuously advanced in the countermeasure. However, in the scenario of multi-unmanned aerial vehicle co-interference, if an effective co-decision mechanism is lacking, conflicts are easily generated among the multi-unmanned aerial vehicles in aspects of interference target selection, interference mode, timing, power allocation and the like. For example, a plurality of unmanned aerial vehicles simultaneously implement high-power interference on the same radar, redundant superposition of interference resources is often caused, the overall interference efficiency of the system is limited, the risks of energy consumption and detection and locking by the enemy radar are obviously increased, and meanwhile, part of high-threat radars can not be effectively inhibited due to unreasonable interference resource allocation, so that the overall survivability of the cluster is reduced. The above phenomenon is commonly referred to in engineering practice as "mutual interference" or "interference resource collision" between multiple drones. In recent years, reinforcement learning and multi-agent reinforcement learning methods are introduced into the field of electronic countermeasure decision-making, but the existing researches are mostly focused on single interference machines or simple cooperative scenes, and the interference resource conflict problem among multiple unmanned aerial vehicles is mostly not systematically solved at the decision level. While the method introduces multi-agent learning, the lack of global constraint on cooperative behavior can still generate redundant interference and resource waste. Therefore, a collaborative decision-making method capable of actively avoiding invalid interference between multiple unmanned aerial vehicles through a learning mechanism in a dynamic electromagnetic environment is needed. Disclosure of Invention In order to solve the problems in the prior art, the invention provides a self-adaptive decision method based on multi-unmanned aerial vehicle cooperative interference, which comprises the steps of constructing a multi-unmanned aerial vehicle cooperative interference task decision model under an electronic countermeasure scene containing a radar and a plurality of unmanned aerial vehicles; the method comprises the steps of carrying out parameterization modeling on interference behaviors of the interfering unmanned aerial vehicles based on a multi-unmanned aerial vehicle cooperative interference task decision model to generate an interference action space, training the multi-unmanned aerial vehicle cooperative interference task decision model based on the interference action space to obtain a cooperative interference decision model, deploying the cooperative interference decision model to each interfering unmanned aerial vehicle, extracting radar signal characteristics and constructing a local observation state by each interfering unmanned aerial vehicle only based on radar radiation signals intercepted by an onboard electronic support measure system, and independently outputting interference decisions under the condition of no centralized control and no explicit communication, thereby completing the cooperative interference task. The invention has the beneficial effects that: Aiming at the problem of dynamic game between multiple unmanned aerial vehicles and multiple radar systems, the invention realizes the cooperative optimization of the multiple unmanned aerial vehicles in the aspects of interference target selection, interference mode, time sequence, power distribution and the like under the condition of not dependin