CN110718098A

CN110718098A - Massive aircraft channel path optimization algorithm

Info

Publication number: CN110718098A
Application number: CN201810749585.5A
Authority: CN
Inventors: 宗鹏
Original assignee: Individual
Current assignee: Individual
Priority date: 2018-07-12
Filing date: 2018-07-12
Publication date: 2020-01-21

Abstract

On the basis of the mass unmanned aerial vehicle identification and channel optimization management technology provided by the invention, the most reasonable flight route is established for the flight of all unmanned aerial vehicles including navigation airplanes according to the theory of the queuing theory. Firstly, designing a route of an airplane to be planned according to a shortest path first rule; and then, according to the number of airplanes on the existing flight path in the system, a queuing model is established, and the congestion probability and queuing time of each node in the path are calculated, so that the whole flight time is calculated. On the basis, optimization adjustment, namely congestion control, is carried out. The concrete measures are as follows: when a new airplane enters a channel or the original balance in operation is damaged, the most reasonable path planning is carried out for all the airplanes according to the cost increment rule on the premise of not generating congestion, the flight time is ensured, the new balance is established, the planning result is transmitted to all the airplanes, and meanwhile, the system flight state database is updated.

Description

Massive aircraft channel path optimization algorithm

The technical field is as follows:

the invention establishes the most reasonable flight route for the flight of all unmanned aerial vehicles including navigation airplanes according to the theory of the queuing theory. Firstly, designing a route of an airplane to be planned according to a shortest path first rule; and then, according to the number of airplanes on the existing flight path in the system, a queuing model is established, and the congestion probability and queuing time of each node in the path are calculated, so that the whole flight time is calculated. On the basis, optimization adjustment, namely congestion control, is carried out. The concrete measures are as follows: when a new airplane enters or the old balance is damaged in operation, the most reasonable path planning is carried out for all the airplanes on the premise of not generating congestion, the flight time is ensured, the new balance is established, the planning result is transmitted to all the airplanes, and meanwhile, the system flight state database is updated.

Background art:

the invention provides a flow management method for aerial three-dimensional channels of mass unmanned aerial vehicles or navigation airplanes based on a queuing theory. On the basis of the mass unmanned aerial vehicle identification and management technology which is proposed previously, the influence of all possible selected flight paths on other flight channels is calculated by a cost increment theory through analyzing the flow state of the whole airspace channel. And then, carrying out integral optimization and informing all unmanned aerial vehicles to carry out flight path adjustment. Effectively avoiding confusion caused by blind flying. At the airspace channel through optimizing, unmanned aerial vehicle and navigation aircraft can fly in order, improve resource utilization.

The invention content is as follows:

including the integration of three aspects of technology. Firstly, the shortest airspace path planning is carried out aiming at the mass aircrafts. And secondly, estimating the flight time of the corresponding path and the congestion probability of each node based on a queuing theory. And thirdly, calculating the influence on other airplanes after selecting the corresponding path based on a cost increment principle, and then performing global path optimization adjustment according to the congestion probability. The specific contents are as follows: the flight path is first selected according to the shortest distance, typically consisting of a number of previously defined channel segments. The congestion probability, as well as the flight and wait times on each channel segment, are then calculated according to queuing theory to determine the actual flight time. And simultaneously, a new jam probability formed after the aircraft enters the channel is used as the basis for system optimization. And finally, carrying out system optimization by adopting a cost increment principle. The method is that the aircraft newly entering the channel will influence the queuing of the aircraft in the original channel during each flight, mainly the congestion probability of the nodes. Therefore, the original airplanes also need to plan the path again, and the planning results can influence the queuing waiting time of the newly entered channel airplane, and need to plan again, so that the congestion probability is reduced to the threshold value, and the steps are repeated until the optimal flight route is found.

Description of the drawings:

FIG. 1 is a schematic diagram of the intersection and critical path of a mass of aircraft channels

FIG. 2 mass aircraft channel optimization flow chart

The specific implementation mode is as follows:

the flight path planning of an aircraft is based primarily on time of flight as a basis for optimization, referred to herein as a cost. Which is represented by the sum of the costs of the individual segmented paths. The time spent by each node by the aircraft newly entering the channel and the time spent by the original aircraft abandoning the original queue due to channel reselection is called cost increment. In the calculation process, the cost increment can be positive and represents the increase of the flight cost of other airplanes in the path; it may also be negative, with the opposite meaning. The affected individual aircraft make up the optimization set, and all paths that intersect with other aircraft are referred to as critical paths, as indicated by the red bold arrow in the center of FIG. 1.

Parameter definition: assuming that the aircraft has k shortest candidate paths, wherein m key paths exist, the change amount of the service of each path to the key link cost is respectively calculated according to a cost increment formula defined below, so as to judge the influence on the aircraft after selecting the route. On the basis, a congestion control mechanism is started, and the purpose of controlling congestion is achieved by comparing congestion thresholds and then redesigning part of routes.

Setting an initial cost L of a certain path₀(i, j) from the time delay of flight E_p(i, j) and node queuing delay E_q(i, j) two parts, expressed as follows:

wherein n is_q(i, j) is the current queue size of the node or the number of waiting planes passing through the node, d_i，jIs the flight distance, v, from node i to node j_i，jIs the flight speed, μ, of the aircraft_i，jIs the node processing capacity, i.e. the number of aircraft that can be processed per unit time, generally related to the geographical environment of flight and to regulatory factors, t_sIs the start time of the calculation cycle.

The critical path cost increment delta (i, j) is the added time delay after the aircraft enters the critical path,

wherein the content of the first and second substances,

for the average waiting time of a new aircraft on the critical path CL,

for the original arrival rate of the aircraft on the path,

for potential aircraft reach, N is the number of nodes that need to be considered. The larger the cost increment of the key path is, the longer the time delay corresponding to queuing is, the higher the possibility of congestion of the path section is, and meanwhile, the larger the influence of the new airplane on the delay of other airplanes selecting the path section is. The core of the optimization is therefore to control the congestion rate.

The threshold value of the congestion probability is defined as

Wherein the content of the first and second substances,

the utilization of a node or the probability of an aircraft selecting entry into the node is determined. When the congestion probability of the path i, j exceeds

Congestion occurs and a control mechanism needs to be started for control.

The algorithm is implemented as follows: the state database in the airspace channel stores the flight state information of all the current airplanes, the algorithm calculates the routes and paths of all the airplanes, ensures the congestion rate of each node or channel path to be kept at the required level, and sends the planning information of the paths or the routes to each airplane. The algorithm flow chart is shown in fig. 2, and the specific steps are as follows:

step 1, respectively obtaining a topological state matrix x and a flight request psi matrix of all flight routes in a current system from a flight state database.

And 2, calculating the blocking probability and threshold value of each path section (node), and judging whether the channel path in the airspace is congested. If so, a congestion control mechanism is initiated. Selecting the airplane to be redesigned into the set psi according to the time required by the flight mission_dAdding a new flyMachine request matrix Ψ_n＝Ψ+Ψ_dUntil the blocking rate of the path or node is less than the blocking threshold

And then removing the congested path from the topology, and avoiding repeated calculation during reselection.

Step 3, calculating initial path cost; is Ψ_nEach aircraft in (1) calculates k_s，dA selectable path is put into the candidate path set K; selecting a set C of critical paths from the above paths_CLThe set of all airplanes passing through the key channel is recorded as L_i，j。

Step 4, calculating the cost L of the critical path_i，jThe incremental costs of the critical path caused by each potential aircraft are arranged in order from small to large.

And 5, sequentially judging whether each route on the critical path is considered by other critical paths. If yes, recording the final route which is the node pair, and deleting the node as other k-1 candidate paths; if not, deleting the route and selecting one of the paths which are not considered to rejoin the critical path. And under the condition that the capacity of the residual channel meets the requirement and no congestion is generated, continuously judging the next route until all routes on the key path are processed and no congestion is generated.

Step 6, from C_CLDelete critical path if C_CLIf not, returning to the step 5, otherwise, entering the step 7.

And 7, recovering the deleted path, returning to the first step, and preparing the next path design, thereby circularly reciprocating.

Claims

1. The application of massive unmanned aerial vehicles and navigation airplanes is an inevitable trend for the development of the future aviation industry. The core problem is the reasonable management of air channels. The invention provides an effective solution based on the queuing theory. The basic method is that firstly, according to the defined air channel, the air route is designed according to the shortest distance, then the statistical model of each route section or node is established according to the flight data of the air plane, thereby calculating the plane queuing number and waiting time of each node and the whole flight time, which are called as the cost. And finally, optimizing the system.

2. The optimization of the system refers to the redesign of routes for the aircraft newly entering the channel and the aircraft already in the channel. The new airplane enters the channel, channel burden is increased, congestion is caused, original balance is damaged, in addition, the original airplane is influenced by various flight factors, planning is changed, channel congestion is caused, and therefore new adjustment is needed. Here, the influence of each aircraft on the channel is quantified and referred to as a cost increment. Entering the channel is positive and changing course and leaving the channel is negative. The cost of each aircraft for various selected paths and the influence of the cost on other aircraft can be calculated according to the cost increment, so that the adjusted cost or route can be calculated.

3. According to claims 1 and 2, the specific operation process is as follows:

the optimization of the air route mainly aims at the maintenance of system balance, and a newly-entered airplane can break the balance, wherein the balance breaking refers to that the congestion rate of a node or a channel exceeds a set threshold to cause corresponding airplane delay. The method mentioned in claims 1 and 2 is used here to plan each relevant route, i.e. the aircraft with staggered overlapping paths, also called critical paths, again, in order to reduce the probability of congestion at the corresponding nodes or routes, and finally to achieve the goal that each aircraft can reduce the time delay as much as possible. The specific process is shown in the attached drawing. Certainly, the system is also volumetric, when the number of the airplanes reaches a certain degree, the optimization algorithm also reaches a limit, and at this time, a priority mode needs to be considered to ensure that the key airplane is preferentially arranged in a channel, and other airplanes can only be abandoned so as to leave the airspace.