Purpose The paper aims to address the tracking algorithm based on deep learning and four deep learning tracking models developed. They compared with each other to prevent collision and to obtain target tracking in autonomous aircraft. Design/methodology/approach First, to follow the visual target, the detection methods were used and then the tracking methods were examined. Here, four models (deep convolutional neural networks (DCNN), deep convolutional neural networks with fine-tuning (DCNNFN), transfer learning with deep convolutional neural network (TLDCNN) and fine-tuning deep convolutional neural network with transfer learning (FNDCNNTL)) were developed. Findings The training time of DCNN took 9 min 33 s, while the accuracy percentage was calculated as 84%. In DCNNFN, the training time of the network was calculated as 4 min 26 s and the accuracy percentage was 91%. The training of TLDCNN) took 34 min and 49 s and the accuracy percentage was calculated as 95%. With FNDCNNTL, the training time of the network was calculated as 34 min 33 s and the accuracy percentage was nearly 100%. Originality/value Compared to the results in the literature ranging from 89.4% to 95.6%, using FNDCNNTL, better results were found in the paper.

Purpose This paper aims to discuss engine health monitoring for unmanned aerial vehicles. It is intended to make consistent predictions about the future status of the engine performance parameters by using their current states. Design/methodology/approach The aim is to minimize risks before they turn into problems. In accordance with these objectives, temporal and financial savings are planned to be achieved by contributing processes such as extending the engine life, preventing early disassembly-reassembly and mechanical wears and reducing the maintenance costs. Based on this point of view, a data-based software is developed in MATLAB (Matrix Laboratory) program for the so-called process. Findings The software is operated for the performance parameters of the turbojet engine that is used in a small unmanned aerial vehicle of Tusas Engine Industry. The obtained results are compared with the real data of the engine. As a result of this comparison, a fault that may occur in the engine can be detected before being determined. Originality/value It is clearly demonstrated that the engine operation in adverse conditions can be prevented. This situation means that the software developed operates successfully.

Purpose The purpose of this study is to detect and reconstruct a fault in pitot probe and static ports, which are components of the air data system in commercial aircrafts, without false alarm and no need for pitot-static measurements. In this way, flight crew will be prevented from flying according to incorrect data and aircraft accidents that may occur will be prevented. Design/methodology/approach Real flight data collected from a local airline was used to design the relevant system. Correlation analysis was performed to select the data related to the airspeed and altitude. Fault detection and reconstruction were carried out by using adaptive neural fuzzy inference system and artificial neural networks, which are machine learning methods. MATLAB software was used for all the calculations. Findings No false alarm was detected when the fault test following the fault modeling was carried out at 0–2 s range by filtering the residual signal. When the fault was detected, fault reconstruction process was initiated so that system output could be achieved according to estimated sensor data. Practical implications The presented alternative analytical redundant airspeed and altitude calculation scheme could be used when the pitot-static system contains any fault condition. Originality/value Instead of using the methods based on hardware redundancy, the authors designed a new system within the scope of this study. Fault situations that may occur in pitot probes and static ports are modeled and different fault scenarios that can be encountered in all flight phases have been examined.

Purpose Fault detection, isolation and reconfiguration of the flight control system is an important problem to obtain healthy flight. This paper aims to propose an integrated approach for aircraft fault-tolerant control. Design/methodology/approach The integrated structure includes a Kalman filter to obtain without noise, a full order observer for sensor fault detection, a GOS (generalized observer scheme) for sensor fault isolation and a fuzzy controller to reconfigure of the healthy sensor. This combination is simulated using the state space model of a lateral flight control system in case of disturbance and under sensor fault scenario. Findings Using a dedicated observer scheme, the detection and time of sensor fault are correct, but the sensor fault isolation is evaluated incorrectly while the faulty sensor is isolated correctly using GOS. The simulation results show that the suggested approach works affectively for sensor faults with disturbance. Originality/value This paper proposes an integrated approach for aircraft fault-tolerant control. Under this framework, three units are designed, one is Kalman filter for filtering and the other is GOS for sensor fault isolation and another is fuzzy logic for reconfiguration. An integrated approach is sensitive to faults that have disturbances. The simulation results show the proposed integrated approach can be used for any linear system.