International Core Journal of Engineering 2020-26 | Page 156

2019 International Conference on Artificial Intelligence and Advanced Manufacturing (AIAM)
Research on the Most Throwing Position of UAV
Based on Mathematical Modeling
Liu Guoliang 1 , Wang Qinjie 2 , Huang Gefan 1 , Tu Lijing 1 , Wu Guodong 1*
1
College of Information and Computer Science, Anhui Agricultural University, Hefei, Anhui, 230036
2
College of Science, Anhui Agricultural University, Hefei, Anhui, 230036
called“DroneGo.”.“DroneGo.” uses drones to transport
supplies, photograph traffic networks and provide
decision-making basis for the Help.Inc Command Center.It
is now necessary to design a transportable disaster response
system in the disaster area to provide three types of
pre-packaged medical supplies for five designated hospitals.
Abstract—In order to help Puerto Rico, our team design a
mathematical model used to develop an efficient scheme. After
analyzing the information, an evaluation mechanism is
established, then analysis is carried out based on a
mathematical programming approach. For the design of the
"Dronego" system, an Fuzzy AHP-based evaluation model is
used to score the performance of each drone under three
criteria of the maximum load, the cost and the capacity of the
drone. The fuzzy weights during calculation have a great
impact on the model. To reduce this effect, the PSO algorithm
is used to correct the weight matrix. The PSO algorithm
searches for the corresponding weight based on the existing
information and modifies the weight matrix, then replaces it
with the model for calculation. Consider the location of the
ISO container by searching a large amount of data, and
summarize four attributes: population, distance from the East
Coast, number of medical items required, and number of
surrounding roads. Through standardizing those four
attributes, the dimension is removed, and the merits and
demerits of the site are evaluated. The first three are selected
as locations. Next, a mathematical programming model is
established by analyzing the ISO container location, taking
into account the system's fastest response and merchant cost
issues. The objective function should be such that the number
of drones in the system and the waiting time of the drug
delivery point are as small as possible. Considering the
objective function under various constraints such as meeting
the needs of the hospital and the estimation of previous
evaluation model for the Drone, a specific DroneGo disaster
response system can be obtained: Drone---H (3), B(3), G(1),
and F(1).Finally,the drone fleet has been designed and the
payload packaging configuration has been obtained for each
drone by the loading capacity of the specific drone and the
requirements of the corresponding location. Then, by
simulating the flight capacity of the drone and the needs of the
transportation destination, the weight of the maximum no-load
distance of the drone are measured, and finally the
transmission route and timetable are obtained.
To discuss the scheme for the DroneGo system, we
analyze the specifications and payload capabilities of the
items . We gather the distributions of the items to design an
evaluation model based on the Fuzzy Analytic Hierarchy
Process (AHP) method instead of calculating directly, which
will cause a large amount of complex calculation. To
simplify the model, we analyze and summarize the
information to produce three distributions: Drone Fleet’s
Max Payload Capability (DFMPC), Drone’s Cost (DC) and
Drone Fleet’s Volume (DFV) as the model’s criteria layer.
To improve the accuracy of our evaluation model, we have
to consider that the relationship during the distributions may
be fuzzy and this will cause the bad performance of our
model. We use particle swarm optimization to correct the
weight in our model.
II. F UZZY A NALYTIC H IERARCHY P ROCESS M ETHOD
This paper assumes that Drone Fleet’s Payload
Capability has the greatest weight, Drone Fleet’s Volume
has the second and Drone’s Cost gets the smallest weight.
Based on the assumptions, we can make a weight matrix
through the method of 1-9 scale for them in Table I.
T ABLE I. O RIGINAL W EIGHT M ATRIX : S CHEME
DC 1 5
DFV
4
DC 1/5 1 1/3
DFV 1/4 3 1
The PSO is used to correct the value and the consistency
check is conducted. We use the ratio of each drone’s max
payload capability to describe the weights of them. But
before starting to calculate, we have to consider how to deal
with the relationship between the drone cargo bays, the
drones, the medical packages and the Standard ISO
Containers. We need to think about how the items are placed
when they are loaded and whether the loaded method fits
the requirements. From Fig.1,we consider that the drug in
the medical package can’t be inverted or tilted, which means
when loaded, we just rotate them in the x-y plane.
Keywords—Analytic Hierarchy Process; Mathematical
Model; Most Throwing Position of UAV; PSO Algorithm
I. I NTRODUCTION
In September, Hurricane landed on Puerto Rico from the
eastern coastline and caused great damage. The entire island
was powered off. In addition, the storm damaged or
destroyed the majority of the island’s cellular
communication networks. Now the Help.Inc organization
needs to establishe a transportable disaster response system
978-1-7281-4691-1/19/$31.00 ©2019 IEEE
DOI 10.1109/AIAM48774.2019.00034
DFMPC DFMPC
134