Development of a 6-DOF Testing Platform for Multirotor Flying Vehicles with Suspended Loads

27 Oct.,2022


WingFlying Drone Thrust Test

1. Introduction

In recent years, multirotor flying vehicles such as quadrotors have gained considerable popularity due to a variety of advantages that they possess, including reliability, simplicity, economy (cheap and easy to manufacture), and agility, among others. This has allowed them to find diverse uses in both civilian and military applications, such as in agriculture, search and rescue missions, reconnaissance missions, inspection (pipeline, risk zone inspections), and aerial mapping. Studies have thus continued to strive to advance this technology.


The development of a multirotor vehicle can be complex, time-consuming [ 1 ], and dangerous [ 2 ]. A typical development process for a multirotor vehicle usually involves modeling, system design, design of the vehicle controller, simulation, and actual flight testing of the vehicle [ 1 3 ]. During the development process, the simulation phase is important, as it can be used to ascertain whether the controllers are functioning correctly. This reduces the risk of dangerous and costly crashes (resulting from faulty controllers). It thus provides a safe and cheaper way for testing controller performance. While simulations serve an important role in the development process, it is known that the performance of the controllers in simulations differs from that in real flight [ 1 ]. This can be due to a variety of reasons, including the inability to completely simulate all flight conditions correctly and the likely existence of errors in any simulation. A solution to this problem is to perform flight tests in a testbed, which is a safe way to test the controllers in more realistic conditions. Most of the published articles in regard to multirotor testbed are related to hardware in the loop (HIL) flight tests. These tests are, however, not considered to be real flight tests, as the sensors that provide the vehicle states are not real [ 2 6 ].


A number of published works have addressed the matter of multirotor testing beds. In [ 7 8 ], the authors reported the development of one-axis test platforms that have been used for control testing, including tuning of the PID controllers. Since these testing platforms enable rotation about a single axis and no translations, they possess limited functionality. Testbeds inspired by the design of the gyroscope have been reported in [ 9 15 ]. These testbeds include 3-DOF designs in [ 9 12 ] that allow for rotation about the pitch, roll, and yaw axes; and 4-DOF designs with the additional DOF of elevation [ 13 15 ]. Six-DOF testbeds for control tests have been developed, such as those documented in [ 1 18 ]. These test beds allow for motion similar to motion in free flight, within some test limits, to avoid crashes. Additional information on these test beds is summarized in Table 1


The test beds reported in [ 1 18 ] have been used for the safe (prevention of crashes and damage and loss to property) development of multirotor vehicles. These testbeds were designed for testing vehicles without attached suspended loads. Vehicles with suspended loads have numerous important applications, such as load delivery, mine detection, and rescue missions. However, the suspended load under such a vehicle creates pendulous motions that have an adverse effect on the performance of the vehicle. These motions need to be damped, and there are a variety of techniques to achieve this. One such method (of interest to this work) is an indirect control approach through cable angle feedback on the load motion to the multirotor vehicle. This technique, generally referred to as cable angle feedback (CAF) control, was pioneered by [ 19 22 ] and has since been used by numerous researchers; see [ 23 26 ]. A search in the literature for test platforms designed for testing vehicles these types of vehicles (vehicles with attached suspended loads) yielded no results. An alternative solution reported in the literature [ 27 30 ] is unrestricted flight (free flight) tests for testing vehicles with suspended loads. These tests included tests for multivehicle collaborative swing load transportation [ 29 ], anti-swing controllers [ 27 30 ], and swing load trajectory tracking [ 28 ] (see Table 1 for summarized details of these tests). While these tests were reported to be successful, there still remained a risk of crashing due to a variety of reasons, such as faulty controllers. There thus appears to be a gap in the literature for testbeds for testing multirotor vehicles with suspended loads. The work in this article sought to address this gap through the development of a testbed for multirotor vehicles with suspended loads. The proposed test platform is expected to help designers to test and optimize new controllers for multirotor flight vehicles in a safe environment. The proposed testbed has good mobility; therefore, it can be used to conduct both indoor and outdoor flight tests. The testbed was also designed to aid in the determination of physical parameters, including the center of mass along the yaw axis, and the moments of inertia about the pitch, yaw, and roll axes, which are usually required in the development of new multirotor vehicles.

Table 1. Information on testbeds for multirotor flying vehicles reported in the literature.

Table 1. Information on testbeds for multirotor flying vehicles reported in the literature.

Number of Degrees of Freedom of the TestbedsSummarized Details of the TestbedType of
Vehicle in the TestReferenceSuspended LoadOne DOFSingle axis test-bench used for PID controller tuningQuadcopter[7]No Suspended loadSingle axis control test bench[8]Three DOFGyroscopic test bench to test for stability tests including over or under actuated UACs as we control loop structuresQuadcopter[9]Simple test rig for testing the designed controller. Rotation about two of the three axes (pitch, raw and yaw) is fixed allowing for rotation on only either the pitch or the roll. Translation is also fixed in all direction except for vertical movement which is also fixable.[10]Gyroscopic 3 DOF test bench for analyzing control systems[11]3DOF test bench to test for modified PID controller for stabilization of a quadcopter. The bed allows for rotation about three axes with translation being restricted.[12]Four DOFTest platform based on gyroscope motion with 3 side circles. Tuning of control parameters. Four axis motion tests including, elevation, pitch, yaw and roll tests.Multi-rotor UAV[13]3DOF test platform where rotation about all axes is allowed whereas translation is restricted, and a 1 DOF test platform complementing the first platformQuadcopter[14]Development of a variable DOF flight control system for a quadcopter.Separate test for pitch, yaw roll and elevation through use of lockable universal joints and roller bearings allowing of up to 4 DOF.[15]Six DOFTest bed to safely test designed UAV. Test platform allows for rotation about the pitch and roll axes and translation along three axes through linear guides.Quadcopter[16]Test platform to test a developed hovering algorithm[17]Testbed designed to evaluate performance of bother attitude and position controllers for multicopter vehicles. Testbed designed to allow 6DOF motion of the multicopter[1]6 DOF test platform to emulate actual free flight to aid in the design and control of UAVs. In the platform the UAV is attached to the end effector of an articulated manipulator[18]Quadcopter test bench for 6 DOF flight controller testing. Utilizes a 6 axes torque-force sensor to simulate the position of the vehicle[3]Unrestricted free flightImplementation of an anti-swing controller on a quadcopter with a suspended load. Test conducted indoor in a cageQuadcopter[27]Suspended loadProposed trajectory tracking controller for a single load carrying quadcopter. Experiments conducted indoor in a cage[28]Development of a collaborative control and transportation of a swing load using multiple multicopters. Experiments conducted indoor[29]Test of controller for a quadcopter with suspended load through window. Test conducted using Astec Hummingbird quadcopter to validate proposed control. Actual test with no testbed[30]