Path planning application algorithm based on parrot algorithm

Abstract

The invention belongs to the field of unmanned driving path planning, and a new algorithm is applied to autonomous trajectory planning of an unmanned vehicle, so that application in the field of path trajectories is expanded, and unique advantages different from those of other traditional algorithms are expected to be obtained. The method comprises the following steps: firstly, carrying out path random point sampling, and carrying out an initialization process; constructing an unmanned vehicle path equation according to a foraging link of a parrot algorithm; unmanned vehicle parking is simulated according to a parking link of the parrot algorithm, wherein the parking link comprises a path distance calculation link; according to the communication link of the parrot algorithm, the traffic jam problem in real life is fitted and solved, so that the unmanned vehicle can achieve the purpose of keeping a distance from other vehicles; according to the fear link of the parrot algorithm, the unmanned vehicle recognizes and avoids large obstacles, and the safety system is maximized. By integrating the steps, the application of the parrot algorithm in autonomous trajectory planning of the unmanned vehicle is integrally completed, so that behavior randomness observed in domestication of the parrot is more properly applied, and path diversity is remarkably enhanced. Meanwhile, by deviating from a traditional exploration-mining two-stage structure, the risk of being trapped in a local optimal value is effectively reduced through the parrot algorithm, meanwhile, the quality of a solution is kept, the parrot algorithm is distinguished from a traditional algorithm through a random structure, and the parrot algorithm is particularly suitable for avoiding the local optimal condition in the path planning problem.

Description

Technical Field Unmanned driving path planning field, specifically a path planning application method involving a Parrot algorithm Technical Background

[0001] Path planning is a key technical link in the navigation and automation system of unmanned vehicles. Its purpose is to guide the mobile unmanned vehicle to move freely in the environment step by step through calculation, avoid obstacles and reach the target location safely. With the development of artificial intelligence, sensor technology and computing power, path planning has been widely used in many fields such as unmanned driving and drones. At present, there are many algorithms for path planning at home and abroad, including traditional algorithms such as A* algorithm and artificial potential field, as well as a series of intelligent optimization algorithms, such as particle swarm algorithm and ant colony algorithm.

[0002] The Parrot Algorithm is an emerging intelligent optimization algorithm inspired by the behavior of parrots. The Parrot Algorithm mainly solves optimization problems by simulating the interaction and collaboration between parrots. Compared with other swarm intelligence algorithms, such as the ant colony algorithm and particle swarm optimization, the Parrot Algorithm has shown good adaptability and efficiency in processing large amounts of data and solving complex problems. It can be applied to various path planning problems, including but not limited to mobile robot navigation, drone route planning, etc.

[0003] The present invention provides a path planning method based on a Parrot algorithm, which is used for performing path planning for an unmanned vehicle.

[0004] The technical solution of the present invention is: a path planning method based on the parrot algorithm, comprising the following steps

[0005] Step 1: Use the grid method to simulate and model the working environment of the unmanned vehicle;

[0006] Step 2: Set the starting point S, the end point E, and the maximum number of iterations Max iter , population size N, fitness function fitness;

[0007] Step 3: Input the starting point location information and the end point location information, and randomly generate a grid map G;

[0008] Step 4: Preprocess the position matrix X according to the generated G to obtain the current global optimal fitness and global optimal position;

[0009] Step 5: Place N parrots at the starting point S and start searching. For each parrot, a probability selection is made for each step. Each step is divided into four behavior choices: foraging, staying, communicating, and fear. After the selection, the position parameters and fitness parameters are updated according to the selected behavior.

[0010] Step 6: Determine whether the parrot has reached the target point E:

[0011] If yes, stop searching and one iteration ends; otherwise, continue searching according to the method in Step 5 until the target point is found;

[0012] Step 7: At the end of an iteration, when all parrots reach the target point E, compare the lengths of all searched paths and update the optimal path GBestX and the optimal fitness GBestF for this iteration.

[0013] Step 8: Determine whether the current number of iterations i has reached the maximum number of iterations Max iter : If yes, output the optimal path; otherwise, go to Step 5 until i = Max iter ; The algorithm ends.

[0014] The foraging behavior function is as follows:

[0015]

[0016]

[0017]

[0018] Where: X i Indicates the current location. Indicates the next step location. represents the average position within the current population, and Levy(dim) represents the Levy distribution, which is used to describe the flight conditions of parrots. best It represents the best position from initialization to the current search, and it also represents the current position of the master. i represents the current iteration number. (X i -X best ) Levy(dim) means to move according to its position relative to the owner. Indicates observing the position of the entire population to further determine the direction of the food. In the Levy distribution, γ is 1.5.

[0019] The dwell behavior function is as follows:

[0020]

[0021] Among them: ones(1,dim) represents the vector of all 1s of dimension dim, X best Levy(dim) indicates the process of flying to the owner, and rand(0,1) ones(1,dim) indicates the process of randomly stopping at a certain part of the owner's body.

[0022] The communication behavior function is as follows:

[0023]

[0024] in: It represents the process of individuals joining the parrot group to communicate. Indicates the process in which individuals fly out immediately after communication. In PO, it is assumed that the probability of these two behaviors is equal, and the average position of the current population is used to symbolize the center of the group.

[0025] The fear behavior function is as follows:

[0026]

[0027] in: Indicates the process of redirecting the flight towards the owner. It indicates the process of moving away from strangers.

[0028] The beneficial effects of the present invention are:

[0029] On the one hand, the present invention introduces the parrot algorithm for path planning, and uses the parrot algorithm to accurately and effectively find a collision-free path from the starting point to the target point, thereby solving the problem of the previous swarm intelligence algorithm falling into the local optimum and the uncertainty of the obtained results. The present invention introduces the Levy distribution to simulate the flight of the parrot, which can more effectively reflect the randomness and suddenness of the parrot during the flight process, which means that the individuals in the population are more likely to cover all parts of the solution space, and will not be too concentrated in a certain local area, and can make full use of the global search ability of the algorithm, helping to find the global optimal solution faster in the solution space and prevent falling into the local optimal solution.

[0030] On the other hand, individuals can perform various behaviors with a certain probability, and this probability can be controlled by adjusting the probability distribution to achieve flexible adjustment of behavior and then realize different algorithmic methods, that is, different population habits. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0032] Figure 1 is a flow chart of the path planning method;

[0033] Figure 2 is the fitness curve diagram;

[0034] Figure 3 This is a grid simulation result diagram of the path planning algorithm based on the Parrot algorithm of the present invention; DETAILED DESCRIPTION

[0035] Embodiment 1: A path planning method based on the parrot algorithm, the main flow chart of which is as follows Figure 1 As shown, the specific steps are as follows:

[0036] Use the grid method to model the working environment. Set the working state space to G, and use the grid method to model the working environment in MATLAB, where the white grid represents the feasible area of ​​the robot, and the black grid represents obstacles, which represent the area that cannot be passed. During the movement of the unmanned vehicle, the obstacles are stationary and the size is fixed. Ignore the size of the unmanned vehicle for now, and use the center point of the grid as the center of mass instead. Establish a Cartesian coordinate system in the grid environment, and one grid represents a position node. Read the size of the generated grid map and assign it to the number of rows xb and the number of columns yb.

[0037] Set the starting point S, the end point E, and the basic parameters of the initial population, including the population size N and the maximum number of iterations Max iter , read the size of the generated grid map and assign them to Xmax and dim respectively. The number of columns in the grid map is the particle dimension, and the number of rows is the range of particle variation.

[0038] According to the grid map, initialize various fitness parameters, including individual optimal fitness PBestF, global optimal fitness GBestF, and global optimal position GBestX.

[0039] Place N parrots at the starting point S and start searching. For each parrot, a probability selection is made for each step. The selection parameter St is assigned with a probability of 1 to 4, that is, each step is divided into four behavior choices: foraging, staying, communicating, and fear. After the selection, the position parameters and fitness parameters are updated according to the selected behavior. The functions of various behaviors are as follows:

[0040] The foraging behavior function is as follows:

[0041]

[0042]

[0043]

[0044] Where: X i Indicates the current location. Indicates the next step location. represents the average position within the current population, and Levy(dim) represents the Levy distribution, which is used to describe the flight conditions of parrots. best It represents the best position from initialization to the current search, and it also represents the current position of the master. i represents the current iteration number. (X i -Xb est) Levy(dim) means to move according to its position relative to the owner. Indicates observing the position of the entire population to further determine the direction of the food. In the Levy distribution, γ is 1.5.

[0045] The stay behavior function is as follows:

[0046]

[0047] Among them: ones(1,dim) represents the vector of all 1s of dimension dim, X best Levy(dim) indicates the process of flying to the owner, and rand(0,1) ones(1,dim) indicates the process of randomly stopping at a certain part of the owner's body.

[0048] The communication behavior function is as follows:

[0049]

[0050] in: It represents the process of individuals joining the parrot group to communicate. Indicates the process in which individuals fly out immediately after communication. In PO, it is assumed that the probability of these two behaviors is equal, and the average position of the current population is used to symbolize the center of the group.

[0051] The fear behavior function is as follows:

[0052]

[0053] in: Indicates the process of redirecting the flight towards the owner. It indicates the process of moving away from strangers.

[0054] After each action, it is necessary to determine whether the updated position information exceeds the boundary value, change the value that exceeds the boundary value to the closest boundary value, and then update the fitness value and the optimal solution; determine the relationship between the fitness of the next position and the individual fitness. If the fitness of the new position is less than the individual fitness, assign the individual fitness to the fitness of the new position and update the position. If the individual fitness is less than the global optimal fitness, assign a value and set the new position as the global optimal position.

[0055] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solution and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A path planning application algorithm based on the Parrot algorithm, characterized in that: The method steps are as follows: Step 1: Use the grid method to simulate and model the working environment of the unmanned vehicle; Step 2: Set the starting point S, the end point E, and the maximum number of iterations Max iter , population size N, fitness function fitness; Step 3: Input the starting point location information and the end point location information, and randomly generate a grid map G; Step 4: Preprocess the position matrix X according to the generated G to obtain the current global optimal fitness and global optimal position; Step 5: Place N parrots at the starting point S and start searching. For each parrot, a probability selection is made for each step. Each step is divided into four behavior choices: foraging, staying, communicating, and fear. After the selection, the position parameters and fitness parameters are updated according to the selected behavior. Step 6: Determine whether the parrot has reached the target point E: If yes, stop searching and one iteration ends; otherwise, continue searching according to the method in Step 5 until the target point is found; Step 7: At the end of an iteration, when all parrots reach the target point E, compare the lengths of all searched paths and update the optimal path GBestX and the optimal fitness GBestF for this iteration. Step 8: Determine whether the current number of iterations i has reached the maximum number of iterations Max iter : If yes, output the optimal path; otherwise, go to Step 5 until i = Max iter ; The algorithm ends.

2. The path planning application algorithm based on the Parrot algorithm according to claim 1 is characterized in that : The foraging behavior function is as follows: Where: X i Indicates the current location. Indicates the next step location. represents the average position within the current population, and Levy(dim) represents the Levy distribution, which is used to describe the flight conditions of parrots. best It represents the best position from initialization to the current search, and it also represents the current position of the master. i represents the current iteration number. (X i -X best ) Levy(dim) means to move according to its position relative to the owner. Indicates observing the position of the entire population to further determine the direction of the food. In the Levy distribution, γ is 1.

5.

3. The path planning application algorithm based on the Parrot algorithm according to claim 1 is characterized in that : The stay behavior function is as follows: Among them: ones(1,dim) represents the vector of all 1s of dimension dim, X best Levy(dim) indicates the process of flying to the owner, and rand(0,1) ones(1,dim) indicates the process of randomly stopping at a certain part of the owner's body.

4. The path planning application algorithm based on the Parrot algorithm according to claim 1 is characterized in that : The communication behavior function is as follows: in: It represents the process of individuals joining the parrot group to communicate. Indicates the process in which individuals fly out immediately after communication. In PO, it is assumed that the probability of these two behaviors is equal, and the average position of the current population is used to symbolize the center of the group.

5. The path planning application algorithm based on the Parrot algorithm according to claim 1 is characterized in that :The fear behavior function is as follows: in: Indicates the process of redirecting the flight towards the owner. It indicates the process of moving away from strangers.

Images