Browse Prior Art Database

PATH PLANNING FOR UNMANNED AERIAL VEHICLE BASED ON ANT CONLONY OPTIMIZATION AND ARTIFICIAL NEURAL NETWORK

IP.com Disclosure Number: IPCOM000244727D
Publication Date: 2016-Jan-06

Publishing Venue

The IP.com Prior Art Database

Abstract

The Path planning for UAV(Unmanned Aerial Vehicle) is the design of a safe, navigable & optimal path for its navigation thereby avoiding obstacle in its path of movement. In this system, the solution for the problem of UAV path planning is viewed as the solution of an optimization problem which could be solved by using a synergy of ACO(Ant Colony Optimization) & ANN(Artificial Neural Network). It is found that the above synergy of ACO & ANN is better than the synergy of GA(Genetic Algorithm) & ANN in terms of performance.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 20% of the total text.

Page 01 of 12

PATH PLANNING FOR UNMANNED AERIAL VEHICLE BASED ON ANT CONLONY OPTIMIZATION AND ARTIFICIAL NEURAL NETWORK

Disclosed is a system which attempts to solve the problem of autonomous path planning for UAV by designing a safe and shortest possible navigable path for it by avoiding its collision with obstacle which comes in its path during its movement from initial to final position. For achieving the above mentioned objective, this system deploys two nature inspired technologies, ACO & ANN. The ACO algorithm is inspired from the foraging behavior of ants and imitates the collision avoidance strategy of the army of ants, when they encounter an obstacle in their path of movement [ 1]. The ACO technique of autonomous path planning for UAV is found to consume less time as compared to GA, and hence its performance can be considered as better than that of GA [2]. Both GA & ACO attempts to find the global optimum solution in the solution space [2], but are found to be slow for implementation in real time situations where time is a crucial issue
[2]. ANNs are good at fitting the given set of input and respective output pairs as a mathematical function in its layers of interconnection quickly, but may converge early towards local optima which is not the best solution for a given optimization problem in the entire solution space [2]. This system uses the synergy of ACO & ANN which results in an ANN trained with the outputs of ACO which is capable of generating the best or global optimum solutions quickly. This system basically works by assuming the problem of path planning as an optimization

problem whose solution could lead to a shortest and safest path for UAV. The solution to this optimization problem of path planning is solved through Ant Colony Optimization (ACO) technique and training the Artificial Neural Network (ANN) from the data set generated from applying ACO over different set of inputs and storing their corresponding outputs. Once the ANN is trained properly it could also help in solving the problem of path planning. The Ant colony optimization technique has been discussed below:

Ant Colony Optimization(ACO) is a kind of Evolutionary Algorithm (EA) introduced by Marco Dorigo & Luca Maria Gambardella in 1997 [1]. ACO exploits the foraging behavior of ants as they always try to find an optimal path between their nest & the food source thereby avoiding in between obstacle naturally, while moving they also deposit pheromone trail along the path. Fig.1 represents the foraging behavior of ants which is used in ACO:

1


Page 02 of 12

Fig.1 How real ants find shortest path.

Initially the ants start travelling from their nest towards food source and arrive at a decision

point as shown Fig.1(a), then some ants choose the upper path while some ants choose lower

path randomly as shown in Fig.1 (b), those ants which had chosen shorter lower path will travel to opposite side quickly than those which had taken longer upper path as all the ants trav...