CN115951714A

CN115951714A - Unmanned aerial vehicle path planning method based on improved pigeon swarm algorithm

Info

Publication number: CN115951714A
Application number: CN202310166206.0A
Authority: CN
Inventors: 代才; 刘柱
Original assignee: Shaanxi Normal University
Current assignee: Shaanxi Normal University
Priority date: 2023-02-24
Filing date: 2023-02-24
Publication date: 2023-04-11

Abstract

The invention relates to an unmanned aerial vehicle path planning method based on an improved pigeon swarm algorithm. In the iterative process, the algorithm can evaluate the fitness value of the pigeons, and the overall objective function value is reduced by updating the speed and the position of the pigeon population, so that a more reasonable flight track is obtained. In the unmanned aerial vehicle path planning experiment, the optimal path searched by the algorithm avoids the problem that the basic pigeon swarm optimization algorithm is easy to fall into local optimization, improves the convergence speed of the algorithm, and balances the convergence and diversity of understanding.

Description

Unmanned aerial vehicle path planning method based on improved pigeon swarm algorithm

Technical Field

The invention belongs to the field of unmanned aerial vehicle path planning, and particularly relates to an unmanned aerial vehicle path planning method based on an improved pigeon swarm algorithm.

Background

In modern society, unmanned aerial vehicles play more and more roles, and practical application scenes of the unmanned aerial vehicles in civil and military use are also more and more in the future. Compared with a manned aircraft, the unmanned aerial vehicle can more conveniently and rapidly reach a specified position to perform tasks, such as battlefield exploration, fire danger exploration and the like. The path planning is to find a specific constrained path of the moving object from the starting position to the target position according to a certain evaluation standard system. The unmanned aerial vehicle path planning needs to comprehensively consider factors such as flight time, flight height and barrier threats, and simultaneously combines the performance constraint of the unmanned aerial vehicle to plan an optimal path from an initial position to a target position.

In real life, there are many problems consisting of multiple conflicting (mutually exclusive) objectives. It is necessary to optimize multiple targets simultaneously and find the optimal solution of the problem under the required conditions, which is a Multi-objective optimization problem (MOP). At present, MOP is widely applied to various fields such as path planning problem, biomedical problem, engineering optimization problem and the like.

Unmanned aerial vehicle path planning is a multi-objective optimization problem. The unmanned aerial vehicle path planning mainly comprises model creation, path constraint, path planning and the like. Common unmanned aerial vehicle path planning algorithms include an a-star algorithm, a Dijkstra algorithm, an artificial potential field method and the like. The appearance of the intelligent optimization algorithm provides a new idea for the unmanned aerial vehicle path optimization problem. Represented by Genetic Algorithm (GA), bee Colony Algorithm (ABC), etc., which are formed by simulating biological, natural, and evolutionary processes, and are global optimization search algorithms based on population iterative evolutionary models. And modifying the population in the solution space through each iteration operation and evolution, and searching by utilizing randomness to obtain a more excellent solution set. The intelligent optimization algorithm can quickly obtain a group of feasible solutions under constraint within the specified time, and the whole search process is shortened, so that the efficiency of operation decision-making is improved. The pigeon swarm Optimization (PIO) is a new heuristic algorithm, which was proposed by professor in stanza in 2012, and has the advantages of easy implementation, relatively fast convergence speed, simple principle, etc., and has been widely applied to engineering practice and other problems. However, the algorithm still has the problems of easy falling into local optimization, low solving speed and the like, and different scholars carry out targeted extension on the pigeon group optimization algorithm in order to improve the performance of the pigeon group optimization algorithm.

Duan et al propose a cooperative control method for close formation of unmanned aerial vehicles based on predation escape pigeon swarm optimization, an outer ring controller is designed based on an artificial potential field method, and the close formation of the unmanned aerial vehicles is converted into an abstract motion in an artificial potential field; an inner ring controller is designed based on a pigeon group optimization algorithm, the control quantity is optimized and solved, the structure of the inner ring controller is adjusted on the basis of following the basic idea of pigeon group optimization, and a predation escape mechanism is introduced to improve the overall performance of the pigeon group optimization algorithm aiming at the problem that the basic pigeon group optimization is easy to fall into local optimization.

Zhang et al adopt a new pigeon swarm optimization algorithm based on Levy flight when optimizing parameters of an active disturbance rejection controller of a small unmanned helicopter. The novel pigeon group optimization algorithm based on the Levy flight improves 2 operators in the basic pigeon group optimization, introduces the search characteristic of the Levy flight into a compass operator, and enlarges the search space of each pigeon by using the Levy flight, thereby overcoming the defect that the basic pigeon group optimization is easy to fall into local optimization to a certain extent; in addition, the Logsig function is used for improving the landmark operator, so that the acceleration of the convergence of the algorithm is promoted, and the comprehensive performance of the algorithm is improved.

Disclosure of Invention

In order to solve the problems that an adopted pigeon swarm algorithm is easy to fall into a local optimal solution and the solving speed is low during unmanned aerial vehicle path planning, the invention provides an unmanned aerial vehicle path planning method based on an improved pigeon swarm algorithm.

The technical scheme adopted by the invention is that the method comprises the following steps:

s1: modeling the flight environment of the unmanned aerial vehicle by adopting a two-dimensional plane space, and representing obstacles with different sizes in a Cartesian coordinate system xoy;

s2: connecting the starting point S and the target point T, dividing the ST segment into D +1 equal parts, making a perpendicular line of the ST segment at each segment point, and making a perpendicular line of each perpendicular line segment L _k Taking 1 discrete point to generate a set of discrete points C, and connecting the starting point S, the target point T and the discrete points C in sequence to form a complete unmanned aerial vehicle flight path;

s3: taking the ST segment as an x axis, and carrying out coordinate transformation on each discrete point to form a new coordinate system and a new discrete point C' set;

s4: setting N pigeons in a path population of a set of discrete points C' to fly to a pigeon nest, wherein the position of each pigeon is a candidate discrete point, and outputting the speed and the position of the pigeon based on an improved pigeon group algorithm;

the specific method for improving the pigeon group algorithm comprises the following steps:

s41: initializing pigeon group positions and related parameters, defining convergence files and diversity files, and randomly generating a group of initial pigeon group speeds;

s42: judging whether the initial pigeon group reaches a stop standard, if so, entering S45, and if not, entering S43;

s43: generating a new pigeon group p according to the initial speed, then simultaneously executing convergence operation and diversity operation on the pigeon group p, determining a convergence file and a diversity file, and taking a union of the convergence file and the diversity file to form an output pigeon group;

s44: judging whether the output pigeon group reaches a stop standard or not, if so, entering S45, and if not, entering S43;

s45: outputting the speed and position of the pigeon;

s5: calculating the fitness of each pigeon in a new coordinate system according to the speed and the position of each pigeon;

s6: and (4) calculating the fitness of each pigeon in the coordinate system, and combining the flight range, the flight height threat and the collision threat of the unmanned aerial vehicle to establish the flight track of the unmanned aerial vehicle.

In S3, the coordinate transformation formula is:

where (((x '(k), y' (k)) denotes the new coordinate, k is an arbitrary value of (1, D), θ is the angle of relative rotation of the original x-axis and the parallel ST line segment, (x) _s ,y _s ) For the original coordinate, the new coordinate x is passed

To obtain, | ST | is the distance from the starting point S to the target point T.

In S43, the convergence operation is performed on the population p as follows:

the method comprises the steps of carrying out grade division on a pigeon group, selecting N excellent solutions from the pigeon group p, then generating a new offspring p1 through crossing of the pigeon group, reserving the solution with the minimum Chebyshev function value in the p1, storing the solution into a convergence file, and determining the convergence file.

The specific manner of the grade division is as follows:

based on the initial number of groups p and a set of generated weight vectors (gamma) ¹ ,γ ² ,…,γ ^m ) And a convergence file, wherein the solution set in the convergence file is graded according to the following formula, and a solution with uniform distribution and better convergence is obtained in an iteration process:

wherein, Z = (Z) ₁ ,Z ₂ ,…,Z _m ) Is a set of ideal points, being the minimum in each object, m is the object dimension, Δ (F (x), γ ⁱ ) Is F (x) -Z and gamma ⁱ The cosine of the angle between them.

The crossing method of the pigeon comprises the following steps:

randomly generating a numerical value J, wherein the J is a random number uniformly distributed between 0 and 1, and when the J is less than 0.5, updating the speed and the position of the pigeon through a compass operator of a pigeon group algorithm;

otherwise, by the formula pop1+ G (X) _g +P _g ) Obtaining the position speed of a new pigeon;

where G is the scaling factor, pop1 is the solution in the population, X _g Global optimal solution, P _g A locally optimal solution.

The pigeon colony algorithm compass operator is as follows:

v for speed and position of pigeon _i ＝(v _i1 ,v _i2 ,…,v _iD ) And X _i ＝(x _i1 ,x _i2 ,…,x _iD ) I =1,2, \ 8230;, N, where N denotes the number of pigeons, and the velocity and position update formula of the pigeons is as follows:

wherein, X _g Is a globally optimal solution, t is the current iteration number, R is [0,1 ]]The compass factor between, rand is [0,1 ]]A random number in between.

The Chebyshev function is used to evaluate the solutions in the convergence profile and to retain the better solutions.

The chebyshev function is as follows:

wherein, γ ⁱ ＝(γ ¹ ,γ ² ,…,γ ^m ) Is a set of weight vectors, Z = (Z) ₁ ,Z ₂ ,…,Z _m ) Is a set of ideal points, the minimum in each target, and m is the target dimension.

In S43, the diversity operation is performed on the population p as follows:

and performing traditional binary crossing on the population p, selecting N better solutions to generate the population p2, selecting any solution with larger distance from the solution p2, storing the solution into the diversity archive, and determining the diversity archive.

In the diversity archive method, the distance calculation formula for any two solutions in the population is as follows:

d(x)＝min{||F(x)-F(y)|| ₂ *||F(x)-F(y)|| _∞ |y∈POP∩y≠x}

S6, establishing an unmanned aerial vehicle flight path expressed as an overall fitness function minimization optimization problem with constraints:

/>

where x is a discrete point selected in the voyage and f ₁ (x) Indicating total flight path, f ₂ (x) Is an index for assessing the threat of flying height, f ₃ (x) Expressed as a collision threat, a whole flight has n small flights, l _i (x) Is the length of flight i _max (x) Is the maximum range of the unmanned plane r _i (x) Represents the threat size, H, due to the flying height of the drone _max Indicating the maximum height, H, of the flight of the drone _min Minimum safe altitude, h, representing the flight of the drone _i Represents the average flying height of the i-th section, H _s Is the height of the terrain, b ₁ And b ₂ Denotes the normal number, w _i (x) Representing the average threat level, p, of the ith track _w Indicates the degree of threat of collision, h _i Indicating the current flying height, H _s Representing the terrain height.

Compared with the prior art, the invention at least has the following beneficial effects: when the path planning of the unmanned aerial vehicle is carried out, the abscissa is equally divided in the coordinate system, the perpendicular line of the abscissa is made at the equally divided point, and the reasonable flight track of the unmanned aerial vehicle can be obtained only by searching the optimal position in the perpendicular direction. In the iterative process, the algorithm can evaluate the fitness value of the pigeons, and the overall objective function value is reduced by updating the speed and the position of the pigeon population, so that a more reasonable flight track is obtained. The unmanned aerial vehicle path planning utilizes an updating strategy of a pigeon swarm optimization algorithm to improve the quality of a solution; new individuals are generated by utilizing a crossover operator to replace inferior individuals, so that the understanding quality is improved; random operators are introduced to improve the diversity of the population; by introducing a selection strategy of double archives, the convergence and diversity of the solution can be balanced. In the unmanned aerial vehicle path planning experiment, the optimal path searched by the algorithm avoids the problem that the basic pigeon swarm optimization algorithm is easy to fall into local optimal, the convergence speed of the algorithm is improved, and the convergence and diversity of the algorithm are balanced and known.

Drawings

In order to more clearly illustrate the technical solution of the embodiment of the present invention, the drawings needed to be used in the embodiment will be briefly described as follows:

FIG. 1 is a flow chart of the present invention based on an improved pigeon flock algorithm;

FIG. 2 is an unmanned aerial vehicle path planning environment model;

fig. 3 is a diagram of a path planning result.

Detailed Description

The invention is further described below with reference to the accompanying drawings:

in order to facilitate the development of the experiment, a plurality of environmental model descriptions are required: assuming that the unmanned aerial vehicle motion space is a two-dimensional plane, in a cartesian coordinate system xoy, the starting point coordinate of the unmanned aerial vehicle is defined as S, the target point is T, a known obstacle exists between S and T, and as shown in fig. 2, a short-flight-path and safe flight trajectory needs to be reasonably planned between S and T in the threat space model. The S-point and T-point are connected first, and then the ST segment is divided into D +1 equal parts. At each segmentation point, make the perpendicular of ST segment, note as (L) ₁ ,L ₂ ,…,L _i ,…,L _D ). At each vertical line segment L _k Take 1 discrete point, resulting in 1 set of discrete points C, which can be expressed as C = { S, L = ₁ (x(1),y(1)),L ₂ (x(2),y(2)),…,L _k (x(k),y(k)),…,L _D (x (D), y (D)), T), connecting the S point, the T point and the discrete points in sequence, thereby forming a complete flight path of the unmanned aerial vehicle.

The invention comprises the following steps:

modeling the flight environment of the unmanned aerial vehicle by adopting a two-dimensional plane space, and representing obstacles with different sizes in a Cartesian coordinate system xoy;

connecting the S point and the T point, dividing the ST segment into D +1 equal parts, making a perpendicular line of the ST segment at each segment point, and making a perpendicular line L at each perpendicular segment _k Taking 1 discrete point, connecting S, T point and these discrete points in sequence to form a complete stripThe unmanned aerial vehicle flight path. In order to improve the convergence rate of the algorithm, the ST segment is taken as an x axis, and each discrete point (x (k), y (k)) is subjected to coordinate transformation, wherein the coordinate transformation formula is as follows:

where (((x '(k), y' (k)) denotes the new coordinate, k is an arbitrary value of (1, D), θ is the angle of relative rotation of the original x-axis to the parallel ST segment, (x) _s ,y _s ) For the original coordinate, the new coordinate x is passed

To obtain, | ST | is the distance from point S to point T. Knowing x (k), y (k), and x ' (k), the y ' (k) coordinate can be found by a coordinate conversion formula, so the set of discrete points C can be simplified as C ' = {0 ₁ (y′(1)),L ₂ (y′(2)),…,L _k (y′(k)),…,L _D (y′(D)),0}；

It is assumed that there are N pigeons flying to the pigeon nest in the path population, the position of each pigeon is a candidate discrete point, and since the x axis is divided into D +1 equal parts, that is, the x coordinate is known, only the optimal position of each ST segment perpendicular line needs to be found. In the iterative process, the algorithm can evaluate the fitness value of the pigeons, and the overall objective function value is reduced by updating the speed and the position of the pigeon population, so that a more reasonable flight track is obtained. As shown in fig. 1, the specific method based on the improved pigeon flock algorithm is as follows: firstly, initializing a population and related parameters, defining a convergence file and a diversity file, and randomly generating a group of initial pigeon speed. Judging whether the initial pigeon population reaches a stopping standard or not, and outputting the speed and the position of the population if the initial pigeon population meets the stopping standard; if not, generating a new population p, and then simultaneously executing convergence operation and diversity operation on the population p. In the convergent operation, the population is graded, N excellent solutions are selected from p, then new offspring p1 are generated by crossing the pigeon population, the solution with the minimum Chebyshev function value in p1 is retained, and the solutions are stored in a convergent file. In diversity operation, a conventional binary crossover is performed on population p, from which N better solutions are selected to generate population p2. And selecting any two solutions with larger distances from the p2 and storing the solutions into the diversity file. Finally, a union set of the convergence file and the diversity file is taken to form an output population and output the speed and the position of the pigeon;

calculating the fitness of each pigeon in a new coordinate system according to the speed and the position of each pigeon;

and establishing the flight track of the unmanned aerial vehicle according to the flight range, the flight height threat and the collision threat.

The method for determining the convergence profile is as follows:

and (3) carrying out grade division on the initial population, then carrying out pigeon crossing, and finally reserving a solution with the minimum Chebyshev function value through a Chebyshev function.

The grade division method comprises the following specific steps:

based on the initial number of groups p and a set of weight vectors (gamma) generated ¹ ,γ ² ,…,γ ^m ) And a convergence file, wherein the solution set in the convergence file can be graded according to the following formula, so that a solution with uniform distribution and better convergence can be obtained in an iteration process:

wherein, Z = (Z) ₁ ,Z ₂ ,…,Z _m ) Is a set of ideal points, being the minimum in each object, m is the object dimension, Δ (F (x), γ ⁱ ) Is F (x) -Z and gamma ⁱ The cosine of the angle between them. The chebyshev function can evaluate the solutions in the convergence profile and retain better solutions.

The chebyshev function is as follows:

the crossing method of the pigeon comprises the following steps:

wherein G is a scaling factor and J is [0,1 ]]With a uniformly distributed random number in between. On one hand, for each solution in the population, the maximum solution of cosine values of included angles between the solution and the uniformly distributed weight vectors is selected as a global optimal solution X through grade division _g And then updating the speed and the position of the pigeons through a compass operator of the pigeon swarm algorithm so as to obtain a new solution. On the other hand, for each solution in the population, any neighbor of the solution is selected as the local optimal solution P _g By pop1+ G (X) _g +P _g ) To obtain a new solution.

The compass operator for the pigeon swarm algorithm is as follows:

v for speed and position of pigeons _i ＝(v _i1 ,v _i2 ,…,v _iD ) And X _i ＝(x _i1 ,x _i2 ,…,x _iD ) I =1,2, \ 8230;, N, where N denotes the number of pigeons, and the velocity and position update formula of the pigeons is as follows:

wherein, X _g Is the global optimum position, t is the current iteration number, R is [0,1]Compass factor between, rand is [0,1 ]]A random number in between.

The method for determining the diversity profile is as follows:

and generating a new group of solutions through the second part of the pigeon group crossing method, and selecting the solutions with larger distance difference for the solutions so as to keep the diversity of the population. The distance calculation formula for any two solutions in the population is as follows:

d(x)＝min{||F(x)-F(y)|| ₂ *||F(x)-F(y)|| _∞ |y∈POP∩y≠x}

wherein | andiF(x)-F(y)|| ₂ And | | | F (x) -F (y) | non-woven phosphor _∞ Representing the difference in distance between any two solutions. The larger the two-norm of two solutions, the greater the probability that the two solutions are retained.

The basic framework based on the improved pigeon swarm algorithm is as follows:

/>

the flight range calculation method comprises the following steps:

f ₁ (x) Representing the total flight path, f, assuming that the drone is flying at a constant speed during the flight ₁ (x) Expressed as:

wherein x is a discrete point selected in the voyage, a whole voyage has n small voyages, l _i (x) Is the length of flight i _max (x) The maximum range of the unmanned aerial vehicle.

The flying height threat calculation method comprises the following steps:

f ₂ (x) Is an index for assessing the threat of flying height, f ₂ (x) Expressed as:

wherein r is _i (x) Represents the size of threat due to the flying height of the unmanned aerial vehicle, H _max Indicating the maximum altitude at which the drone is flying, H _min Minimum safe altitude, h, representing the flight of the drone _i Denotes the average flying height of the i-th section, H _s Is the height of the terrain, b ₁ And b ₂ Indicating a normal number.

The collision threat calculation method is as follows:

wherein w _i (x) Representing the average threat level, p, of the ith track _w Indicates the degree of threat of collision, h _i Indicates the current flying height, H _s Representing the terrain height.

To sum up, the unmanned aerial vehicle path planning can be expressed as a constrained global fitness function minimization optimization problem:

simulation experiment and result:

in order to verify the effectiveness of the algorithm, a simulation experiment was performed using a standard particle swarm algorithm and a standard pigeon swarm algorithm, respectively. The experimental environment is WINDOWS 10; MATLAB R2014a; a memory 16GB; particle swarm learning factor c ₁ ＝2，c ₂ =2, inertial weight w =0.7, randomly generating a set of [ -0.5,0.5]The numerical value between the two is used as the pigeon initial speed of the algorithm and the pigeon group algorithm provided by the invention; the maximum iteration number K =500, and the population size N =50; table 1 shows the spatial threat distribution, with starting point coordinates (100, 400) and target point coordinates (800, 900); fig. 3 shows unmanned aerial vehicle path planning results for three algorithms; obstacles are represented by black circles; solid lineThe dotted line and the dotted line respectively represent the optimal flight path of the unmanned aerial vehicle searched by the algorithm, the particle swarm algorithm and the pigeon swarm algorithm. As can be seen from fig. 3, the algorithm proposed herein draws a smoother and more reasonable path with better obstacle avoidance capability, and verifies the superiority of the algorithm proposed herein.

TABLE 1 spatial threat settings

In order to further verify the effectiveness of the algorithm, an ultra volume evaluation index (HV) is selected as an index for measuring the performance of the algorithm in the algorithm comparison experiment. The hyper-volume is used for describing the evaluation of the performance of the multi-objective optimization algorithm by calculating the value of the hyper-volume of all non-dominant solutions and a selected reference point enclosing a space. The larger the value of the over volume, the better the performance of the algorithm. On the one hand, there are two ways to select the reference point, namely, the worst point on each dimension of the non-dominated solution set and the worst point in a loose form. It can be seen that only the reference point needs to be selected to evaluate the solution set performance. It can also be seen that the choice of reference point has a very large effect on the value of the hyper-volume. On the other hand, the time taken to calculate the hyper-volume value each time is very large. It is mathematically defined as follows:

s represents a pareto optimal solution set, λ represents a Leber-Bege measure, v _i Represents the hyper-volume value of the reference point and each non-dominated solution. In order to avoid reducing errors caused by experiments, required input variables in simulation experiments are kept consistent. The following table shows HV's obtained by solving the problem of unmanned aerial vehicle path planning using the algorithm, pigeon swarm algorithm and particle swarm algorithm presented hereinThe value is obtained.

TABLE 2 HV mean and standard deviation calculated by the algorithm, pigeon flock algorithm and particle swarm algorithm provided by the invention

Algorithm	HV
		The algorithm proposed by the invention	4.5463+05(4.43-04)
Pigeon group algorithm	4.2167+05(1.64e-03)
		Particle swarm algorithm	3.6528e+05(1.58e-01)

As can be seen from Table 2, the HV value of the algorithm provided by the invention is the largest, the next to the HV value of the pigeon group algorithm, and the HV value of the particle group algorithm is the smallest. Therefore, the overall performance of the algorithm provided by the invention is superior to that of a standard pigeon swarm algorithm and a particle swarm algorithm, and the algorithm provided by the invention can effectively solve the problem of unmanned aerial vehicle path planning.

Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can make modifications and equivalents to the embodiments of the present invention without departing from the spirit and scope of the present invention, which is set forth in the claims of the present application.

Claims

1. An unmanned aerial vehicle path planning method based on an improved pigeon swarm algorithm is characterized by comprising the following steps:

s2: connecting the starting point S and the target point T, dividing the ST segment into D +1 equal parts, drawing a vertical line of the ST segment at each segmentation point, and drawing a vertical line L at each vertical line _k Taking 1 discrete point to generate a set of discrete points C, and connecting the starting point S, the target point T and the discrete points C in sequence to form a complete unmanned aerial vehicle flight path;

s41: initializing pigeon group positions and related parameters, defining a convergence file and a diversity file, and randomly generating a group of initial pigeon group speeds;

s44: judging whether the output pigeon group reaches a stop standard, if so, entering S45, and if not, entering S43;

s45: outputting the speed and position of the pigeon;

2. The unmanned aerial vehicle path planning method based on the improved pigeon swarm algorithm as claimed in claim 1, wherein in S3, a coordinate transformation formula is as follows:

where (((x '(k), y' (k)) denotes the new coordinate, k is an arbitrary value of (1, D), θ is the angle of relative rotation of the original x-axis to the parallel ST segment, (x) _s ，y _s ) For the original coordinate, the new coordinate x passes

3. The method for unmanned aerial vehicle path planning based on the improved pigeon swarm algorithm according to claim 1, wherein in S43, the convergence operation on the swarm p is as follows:

4. The unmanned aerial vehicle path planning method based on the improved pigeon swarm algorithm according to claim 3, wherein the specific manner of grading is as follows:

based on the initial number of groups p and a set of generated weight vectors (gamma) ¹ ，γ ² ，...，γ ^m ) And a convergence file, wherein the solution set in the convergence file is graded according to the following formula, and a solution with uniform distribution and better convergence is obtained in an iteration process:

wherein, Z = (Z) ₁ ，Z ₂ ，...，Z _m ) Is a set of ideal points, minimum in each object, m is the object dimension, Δ (F (x), γ ⁱ ) Is F (x) -Z and gamma ⁱ The cosine of the angle between them.

5. The unmanned aerial vehicle path planning method based on the improved pigeon swarm algorithm according to claim 3, wherein the pigeon swarm crossing method is as follows:

6. The improved pigeon loft algorithm based unmanned aerial vehicle path planning method of claim 3, wherein the pigeon loft algorithm compass operator is as follows:

v for speed and position of pigeon _i ＝(v _i1 ，v _i2 ，...，v _iD ) And X _i ＝(x _i1 ，x _i2 ，...，x _iD ) Express, i =1, 2., N, where N denotes the number of pigeons, the velocity and position update formula for the pigeons is as follows:

wherein X _g Is a globally optimal solution, t is the current iteration number, R is [0,1 ]]The compass factor between, rand is [0,1 ]]A random number in between.

7. The improved pigeon loft algorithm based unmanned aerial vehicle path planning method of claim 3, wherein a Chebyshev function is used to evaluate solutions in the convergence file and to retain better solutions;

the chebyshev function is as follows:

wherein, gamma is ⁱ ＝(γ ¹ ，γ ² ，...，γ ^m ) Is a set of weight vectors, Z = (Z) ₁ ，Z ₂ ，...，Z _m ) Is a set of ideal points, the minimum in each target, and m is the target dimension.

8. The unmanned aerial vehicle path planning method based on the improved pigeon flock algorithm according to claim 1, wherein in S43, diversity operation is performed on the flock p as follows:

9. The unmanned aerial vehicle path planning method based on the improved pigeon loft algorithm as claimed in claim 8, wherein in the diversity archive method, the distance calculation formula for any two solutions in the population is as follows:

d(x)＝min{||F(x)-F(y)|| ₂ *||F(x)-F(y)|| _∞ |y∈POP∩y≠x}

10. The method for unmanned aerial vehicle path planning based on the improved pigeon swarm algorithm according to claim 1, wherein in S6, an unmanned aerial vehicle flight trajectory is established to represent an overall fitness function minimization optimization problem with constraints:

/>

where x is a discrete point selected in the voyage and f ₁ (x) Indicating total flight path, f ₂ (x) Is an index for assessing the threat of flying height, f ₃ (x) Expressed as a collision threat, a full flight has n small flights, l _i (x) Is the length of flight i _max (x) Is the maximum range of the unmanned plane r _i (x) Represents the size of threat due to the flying height of the unmanned aerial vehicle, H _max Indicating the maximum height, H, of the flight of the drone _min Minimum safe altitude, h, representing the flight of the drone _i Represents the average flying height of the i-th section, H _s Is the height of the terrain, b ₁ And b ₂ Denotes the normal number, w _i (x) Representing the average threat level, p, of the ith track _w Indicates the degree of threat of collision, h _i Indicating the current flying height, H _s Representing the terrain height.