introduce
Path planning refers to finding routes and paths that meet certain performance indicators and certain constraints between the target point and the starting point for moving objects. At present, the research on route planning is mainly used for the flight route selection of missiles, torpedoes, airplanes and other aircraft, but the systematic research on reconnaissance route of unmanned aerial vehicles is rare. In reference [3], although the ant colony algorithm is also used for path planning, the influence of the existence of threat points and the value of target points on the path is not fully considered, and the ant colony algorithm is not innovated in terms of heuristic factors and initial intensity of pheromones. In foreign related literature, due to the long range of U.S. UAV, there are relatively few constraints on its route planning, and the content available for reference is limited. However, the research on the special combat style of island attack campaign is still blank. Based on this background, this paper studies this problem in order to give full play to the maximum operational effectiveness of UAV and reduce the damage probability of UAV as much as possible.
1. Analysis of factors affecting route planning
There are four main factors that affect the route planning of reconnaissance UAV.
1. 1 target value
The target value is a comprehensive index (expressed by Vm) to measure the necessity of fire assault on the target at a certain moment. The value Vm of each target can be obtained by analytic hierarchy process, or the relative value coefficient Ku of each target can be obtained by normalization, thus measuring the importance of the target.
When scouting different targets, the effective reconnaissance time can be arranged for the target with higher value, while the effective reconnaissance time should be appropriately compressed for the target with relatively low value.
1.2 Effective flight time (distance)
The main purpose of reconnaissance is to find valuable targets and describe their status in time, so the probability of finding targets is an important indicator to measure whether the route is reasonable or not. The closer to the target, the longer the reconnaissance equipment on the plane will search the target area, and the greater the probability of finding the target.
In order to obtain the effective information of the target, the UAV must approach the target and keep the target within the range of its airborne electronic and optical reconnaissance equipment. In order to monitor a target in real time, reconnaissance UAV must hover over the target, so that the target is under the surveillance of airborne equipment for a long time. Therefore, the probability of finding the target can be characterized by the effective flight time. It represents the total reconnaissance and monitoring time of the reconnaissance drone to the target. For the convenience of handling, if the reconnaissance UAV flies at the same speed, its effective reconnaissance flight time can also be converted into effective flight distance representation.
1.3 viability
Reconnaissance UAV must have certain survivability to complete reconnaissance mission. Its survivability is mainly related to the stealth avoidance performance of reconnaissance UAV, the performance of enemy radar and air defense weapons. In other words, the viability of reconnaissance UAV is not only affected by its own fragility, vulnerability and reliability, but also by the enemy's reconnaissance, detection and strike capabilities.
Judging from the process of reconnaissance UAV completing its mission, it includes three processes: launch, normal flight and breakthrough interception. If probability Pf, Pl and Ps are used to represent the completion of the three processes.
1.4 range (fuel quantity) limit
Voyage refers to the maximum horizontal distance that a reconnaissance drone can fly without refueling after taking off, that is, the flight distance. It is an index to characterize the long-distance and lasting flight capability of reconnaissance UAV. Because the amount of refueling on the ground is limited at one time, its route is bound to be limited by the voyage, and because of the limited range of radio, the position where the aircraft performs its mission cannot exceed its operational radius.
2. Route planning modeling
Reconnaissance drones usually perform specific reconnaissance and surveillance tasks. The commander's expected goal is to find as many targets as possible within a limited flight time and range, and at the same time pay the least price.
As far as the constraint conditions of route planning are concerned, firstly, the threat amount should not exceed the commander's permission range, and secondly, the total flight distance of reconnaissance UAV should not exceed the range of reconnaissance UAV. Once one of the two cannot be established, it means that the required task cannot be completed, that is,
3. Ant colony algorithm and its improvement.
Ant colony algorithm, as a new computing mode, is introduced into the field of artificial intelligence, which is called ant system. The system is based on the following assumptions:
(1) Ants communicate through the environment. Each ant only responds according to the surrounding local environment and only affects the surrounding local environment;
(2) The ant's response to the environment is determined by its internal model;
(3) At the individual level, each ant only makes an independent choice according to the environment. At the group level, the behavior of a single ant is random, but the ant colony forms a highly orderly group behavior through self-organization process.
3. 1 Route planning characteristics based on ant colony algorithm
The route planning method of reconnaissance UAV based on ant colony algorithm can ensure that a flight route with low detectable probability and acceptable range can be obtained when formulating the route. This route planning method also has the following characteristics:
(1) New information will be quickly added to the environment under the reinforcement of the continuous spread of bioinformatics hormones by ants, while the old information will be continuously lost due to the evaporation and renewal of bioinformatics hormones, showing dynamic characteristics;
(2) Searching for the optimal path through the cooperation of many ants has become the path chosen by most ants, which is synergistic;
(3) Because many ants feel the scattered bioinformatics hormones in the environment, they also emit bioinformatics hormones themselves, which makes different ants have different selection strategies and are distributed. These characteristics are consistent with many requirements of the future battlefield, so it is feasible and forward-looking to use ant colony algorithm to plan the route of reconnaissance UAV.
3.2 Improvement of Ant Colony Algorithm
(1) the initial value of ij (t)
In order to better consider the threat, the trajectory strength is defined under the initial conditions, and the path with high trajectory strength is optimized according to the ant's path selection, while the path planning of UAV should better choose the path far from the threat point. Then it can be defined that the initial intensity of the trajectory is inversely proportional to the distance. That is, the closer the route is to the threat point, the smaller the pheromone intensity. For each path between two target points, the initial intensity of pheromone trajectory.
4. The realization of route planning of unmanned reconnaissance aircraft based on improved ant colony algorithm.
4. 1 Initial conditions of route planning
Ant colony algorithm is mainly used for multi-target search and reconnaissance route planning, that is, route planning needs to get the number and order of each target, so that reconnaissance UAV can pass as many target points as possible.
In the initial planning process, in order to facilitate the implementation of ant colony algorithm, the coordinate system is determined first, and the above-mentioned target points and threat points are represented by the coordinate system, which is convenient for practical operation.
Suppose that in the battle to seize the island, a city is taken as the coordinate point (100, 100), and a plane rectangular coordinate system of 1 coordinate system with a unit length of 3 kilometers is established (this is fully considering that all major valuable points are included in (120×1). Then you can determine the coordinate system position of the above points and get the coordinates of each point. At the same time, the value coefficient of each target point can be obtained by AHP (the specific process is omitted).
4.2 Implementation of Ant Colony Algorithm Model
Determination of initial parameters of ant peritrophic system
For the convenience of calculaTion and representation, the target point is defined as vector Mi (where I = 1, 2,3, …, 12) and the threat point is defined as vector ti (where I = 1, 2,3). Ant colony algorithm is used to solve the traveling salesman problem at the target point. The ant colony algorithm developed at present includes ant density system, ant quantity system and ant perimeter system, but the latter is mostly used in practical application. In order to simulate the behavior of ants in the system conveniently, labels are defined.
4.3 Ant Colony Algorithm Model Analysis
By comparison, the objective function value and route planning diagram in various situations are qualitatively analyzed. It is not difficult to find that considering the value of the target point and the threat of the threat point, the route should avoid the threat as much as possible and give priority to the point with higher target value. In this way, the damage probability of the UAV is low, and if a damage event occurs, the overall value of the discovered target is the greatest.
Quantitative analysis is carried out for four situations, assuming that the commander's tendency is 0.6, that is, the threat cost is slightly considered. 2000 meters means that the effective reconnaissance distance of each target is 2000 meters, and the target function value is calculated. It can be seen that although the total length of the route is the largest, the overall objective function value is also the largest, and the range is the best, that is, the reconnaissance drone must pass through these target points in turn.
5. Concluding remarks
Through the above analysis, under the given reconnaissance task of reconnaissance UAV, the optimal initial route can be obtained by operation, which can effectively improve the reconnaissance efficiency of UAV and reduce the damage probability of UAV, and has certain guiding significance for how to use reconnaissance UAV in the current military struggle preparation. With the improvement of the performance of our reconnaissance UAV and the continuous enrichment of its models, how to plan the routes of these models in the future island attack campaign needs further discussion.