System THEORY | UAV for Disaster Surveillance | | | Arun Bhatta | S0254398 |

|

Contents Executive Summary 1 Introduction 2 System Test, Evaluation and Validation 2 Validation 3 Human Factor 3 Optimization in design and operation for reliability and maintainability 4 Reliability Assessment 5 Conclusion and Recommendation 5 Bibliography 7

Executive Summary
There are several UAVs which are widely used in civil and military application. The continuous use of UAV has demanded a reliable and low cost UAV system. Since its use in increment it has undoubtedly raised the question about the reliability of these systems. Thus, requiring improvement in the modeling, testing and flight control for the small UAV’s. This reports aims to provide an integrated framework with systematic, schematic procedure so that the UAV can be synthesized and validate flight controllers. This approach will help FAA and EASA certification of UAV system very rapidly. The first part of this report provides a brief introduction regarding the failure of the components. The second part discusses about the System Test and validation of the UAV. The last part of the report analyzes the human factor issues involved in the UAV.

Introduction
The UAV is not a new concept as it was in the situation of First World War. The roles and responsibilities that it carries has increased its utilization in the field of defense, disaster surveillance etc. Increment in utilization has accompanied by more number of failures rates compared to commercial and conventional airframes. There is about 1 per 100,000 flight hours’ general aviation mishap which includes UAV and about 1 per 1000 flight hours that includes military UAV. If these mishaps is to be calculated the failure rates of UAV is about 25% [ (Tvaryanas, Thompson, & Constble) ]. Despite of failures many forms of UAV’s are being used in patrolling of borders, firefighting, disaster relief, search and rescue missions, training, research and development . Currently, UAV are used for commercial excitement such as delivery of packages and logistics. Failure in these kind of system, which are not expected should be rigorously tested or should be well manufactured that may possess a potential for danger and safety risk [ (Ballenger, 2013) ].
The rate of failure rate is determined to have a significant increment as they shall be developed from redundant systems like sensors and wireless communication skills. Therefore, the reliability of these systems should be more obvious as they have become more integrated into our commercial airspace. The UAV can perform dangerous missions such as firefighting, and is able to perform its operation continuously without putting the safety of human operators at risk. As mentioned previously, the delivery of logistics through UAV has significantly decreased the investments for labor and delivery trucks. Ironically, this leads to perception of UAV as consumable hardware which may increase the risk and these activities might lead to gains in reliability and safety.
Although there is no significant record of any casualties due to UAV mishaps, these UAV’s are not yet certified to fly over European skies despite of getting approval from FAA in United States. So, what is the reason besides there use is so rare despite of being conspicuously advantageous? One of the reasons undoubtedly is an insufficient level of reliability of current solutions that has leaded to a potentially unacceptably high probability of accident or a catastrophe [ (Krawczyk, 2013) ].
System Test, Evaluation and Validation
The UAV is first designed in CATIA V5, then after completing its preliminary stage it is further tested aerodynamically, structurally, etc. Then the aircraft is tested by Protean Levita Method which includes design and analysis of control system, aerodynamics, and structural test of the UAV. In addition to that, the UAV is further tested for GSM control UAV in which it was tested for long range communication with UAV for its controls under manufacturing and testing flight condition. After that aircraft is tested for Aircraft Intelligence Requirement. In this test the mother board of the UAV is tested for independent decision making for defined tasks under any circumstances. The aerodynamic analysis of the UAV analyzes the flow properties around the body of the UAV which can be achieved by software’s like ANSYS. For the dynamic analysis of the UAV, MATLAB is used as virtual model of control mechanism to check the control and the response of the UAV. The structural analysis of the UAV’s ribs and spars located in the wing are analyzed using NASTRAN PATRAN, which is structural analysis software.
Validation
Numerical Model analysis of UAV system
The major objective of this simulation was to create a MATLAB model to describe and represent the dynamics of an actual UAV platform within MATLAB was used to develop the system and sub-systems of the model. The mathematical model was developed by assuming the individual components of the UAV to have a linear behavior. The control loops for altitude, roll, pitch, and yaw were coupled together and the control loops for X- and Y-positions were coupled together. PID controller was used for altitude control while PD controller was used for roll, pitch, yaw, X-position and Y-position control. The error ratio observed for position while only using the attitude controller was found to be a maximum of 0.0025. This ratio was reduced to a negligible factor by using the position controller but at the expense of a 1 second lag between the intended time and the actual time for the UAV to reach a required position. The use of a mathematical model to improve the behavior of a UAV was highlighted. By following a similar procedure, the stability and the behavior of a UAV intended for other purposes too can be improved.
The analysis of the results obtained from various simulations performed in this project helps us conclude that the use of position controller effectively improves the accuracy and responsiveness of the UAV compared to the use of only an attitude controller. Improvements in error ratio reduced from 0.0025 to negligibly small amounts and decrease of the time lag from about 5-10 seconds down to 10 seconds highlight this fact. The nature of the altitude controller in regulating the height of the UAV helps us understand that this model is more suitable when height is increased in 10 meter steps and decreased in 5 meter steps. This process, despite being tedious to a recreational user, is more beneficial in areas of disaster management where the UAV is to reach the desired altitude in the quickest time which is evident when the height is stepped. The presence of a steady state lag of 1 second in the behavior of the UAV to follow a specific path is not noticeable when the UAV is expected to cover distances in the range of one locality or one village. This is, however, an undesirable effect. But the fact that the error ratio of the UAV while following a path was reduced to practically nil over the course of local UAV operation outweighed the previous undesirable effect. The use of the mathematical model to improve the behavior of hovercrafts or any system in question is evident due to this project. A user wanting to design his own UAV system can easily do so following the same path and methodology followed by us. With ever increasing future application areas and vast potential of UAV in several fields, we hope this project will provide a baseline for designing and analyzing their control systems and improving their effectiveness
Human Factor
In the development of aviation throughout these years, the most significant error for safe and reliable operation is human error. Although, these machines don’t have pilot on board, some operation experience shows that there is a hazard presented by humans. May be at some stage in future the pilots on ground could be replaces by some kind of autonomous flight system. As mentioned earlier the accident rate for UAV is much higher than that of the other commercial aircrafts. According to Johnson, the loss of unmanned aircraft is much higher than the manned, he also concluded that human error was one of the factors in the loss [ (Johnson, 2003) ]. This is due to the operators of UAV have don’t have real time information and have few methods of fault recovery.
Depending on the type of UAV 2-17 % of the UAV accidents are reportedly to be caused by Maintenance Factors [ (Williams, 2004) ]. His reported concluded most of these accidents were cause of Electromechanical Failure or material failure in conjunction with other failure which was about 45% of the error than that of the operator error which was about 32%. Later on, it was found that large proportion accidents involved Human Factors. The studies done by these researchers found that there might be a greater of system reliability to UAV’s than compared to the conventional aircraft.
One of time consuming and critical activities in aviation is maintenance. According to recent estimation, 12- Man hour of maintenance error occurs for every hour of flight. Any activity that is performed on aerial vehicle for its safe operations before or after the flight is defined as maintenance. It mainly includes assembly, disassembly, fueling, pre-flight inspection, repairs and software updates. Further maintenance can be classified as preventive and corrective maintenance. Non routine or unscheduled task such as hard landing, lightning strike are classified as corrective whereas preventive maintenance are done before problem occurs such as routine task and inspection which may require limited range of skills and knowledge.
Each disturbance of an otherwise functioning system for maintenance introduces the risk of a maintenance induced failure [ (Kletz, 2001) ]. The most common cause of failure of system is deficient of maintenance which is recognized in a range of maintenance. A tradeoff between this risk and the expected benefits of the maintenance procedure is involved in preventive maintenance. Maintenance related system error can be decreased to ensure that preventive maintenance is adopted if any information of component history is available.
Most of the accidents caused in aviation about 70% is due to pilot error whereas 15% of the accidents occur due to deficient of maintenance and inspection. Instead of eliminating the human error, the removal of on board pilot in UAV has accentuated the importance of ground based support activities for UAV operation that has shifted the risk of human error to maintenance personnel. Human factors on UAV has a range of issues that is related to automation, control and interface issues, air traffic management, and qualification issues for UAV operators but there is no studies have been found typically studying the human factors on UAV maintenance [ (McCarley & Wickens, 2005) ].
Optimization in design and operation for reliability and maintainability
Most of the failures in UAV are electromechanical error. However, in case of military UAV it has increased protection against the failure induced by the humans by improving the software’s related flight control. Moreover, it has incorporated multiple sensors so that it can detect and predict the rate of failure. However, the low cost commercial system has attracted some of the leaner and cost sensitive sized business but don’t have the same level of redundancy of reliability as of military. So, electronic component of such machines will have greater reliance on the reliability. There might be a dramatic increment in UAV mishaps due to share of electronic failures.
As it can be seen from my previous assessment UAV system uses Flight Computer, telemetry radio, wireless modem, Failsafe switch, A/D Converter, sensors, batteries, GPS unit and antenna and Pulse Width modulation (PWM). So these kind of products has several chances of fails. The cause of failure of these assemblies might occur due to ESD ( Electro Static Discharge), Electrical Overstress ( EOS) from a power surge, vibration causing connector and solder joint issues along with thermal fatigue of the interconnects.
Reliability Assessment
To reduce the amount of component failure the designers of UAV have been using many Non-Electronic Parts Reliability (NPRD)-2011, Telcordia SR332. After usage of this product, it was found that failure rate of the commercial product was much higher than those of the military grade. So to do study further, Physics of Failure (PoF) approach was required.
The key question that is to be asked are: The environmental ranges that UAV’s can withstand (For e.g. temperatures, pressure, vibration caused by rotor, shock on landing, turbulence) without any electrical failure. To understand why failure occurs, it is very important to learn failure analysis at component and material level so that those failures could be mitigated in the future.
The state of environment that UAV might face include temperature range from -40 C to 70 C, vibration, shock, moisture, pressure and storage. To avoid the issues with crashes and personal safety, it would be beneficial if these stresses could be understood earlier. Keeping this in mind, there are two things that can provide a greater level of reliability assurance: Critical Component Testing and Physics of failure based design analysis. Maximum of the UAV are dependent on wireless transmission for data transfer. On the other hand, the robust series of overlapping communication system has been adopted by militaries UAV to ensure continuous transmission during its operation.

Conclusion and Recommendation
This report assumes that first order equations are sufficient to describe the components of the UAV system. Future researchers can implement a model based on second order equations to minimize the errors in the response of individual components and for better accuracy. The control loops for the attitude values and altitude along with the horizontal positions are coupled in this model. These can be decoupled, which will prevent the transmission of error from one state to another and facilitate independent operation of each controller. The use trial and error method of PID tuning is not recommended as there may be errors on the three gain constants of the PID controller. The model can be improved by considering a way to implement adverse weather conditions in the external disturbances block to obtain stable flight characteristics even in windy and rainy conditions. An inherent drawback of this model is that there is a misrepresentation in conditions involving startup of UAV in elevated pedestals. A technique can be implemented which resets the initial conditions to zero when the UAV is at rest on elevated pedestal. Moreover, this project is only a part of complete UAVdesign, and the project still has remaining work to be verified by implementing it in a custom controller in an actual UAV.

Bibliography
Ballenger, K. (2013). Unmanned Aircraft System- General Overview. Washington: AIAA.
Johnson, E. (2003, September). Making UAVs smarter. USA: Aerospace America.
Kletz, T. (2001). An engineer’s view of human error. London, U.K: Taylor and Francis.
Krawczyk, M. (2013). Conditions for Unmanned Aircraft Reliabilitty Determination. Maintenance and Reliability , 15 (1), 31-36.
McCarley, J., & Wickens, C. (2005). Human factors implications of UAVs in the National Airspace. New Jersey: FAA.
Tvaryanas, M. A., Thompson, W., & Constble, W. The US military UAV Experience: Evidence-Based Human System Integration Lessons Learned. Strategies to Maintain Combat Readiness during extended deployment- A human system approach (pp. 5.1-5.24). Neuilly-sur-Seine: RTO-NATO.
Williams, K. (2004). A summary of unmanned aircraft accident/incident data: Human Factors Implications. U.S Department of Transportation, FAA, Office of Aeropsace Medecine.
Weisstein, E. W., 2012. Euler Angles - Wolfram MathWorld. [Online] Available at: http://mathworld.wolfram.com/EulerAngles.html [Accessed 15 January 2015]. Wikimedia Commons, 2008. Berguet Gyroplane 1902. [Online] Available at: http://en.wikipedia.org/wiki/File:Breguet_Gyroplane_1907.jpg [Accessed 15 January 2015]. Wikimedia, 1924. Wikipedia Commons. [Online] Available at: http://en.wikipedia.org/wiki/File:Oemichen2.jpg [Accessed 13 June 2015]. Zhao, W. & Go, T. H., 2014. Quadcopter formation flight control combining MPC and robust feedback linearization. Journal of the Franklin Institue, 351(1), pp. 1335-1355.