-
Notifications
You must be signed in to change notification settings - Fork 426
Interesting JSBSim Examples
Testing JSBSim's trim routines to confirm that an aircraft can be trimmed for inverted flight for a range of airspeeds.
Using JSBSim to demonstrate the aerodynamic effect of pitch rate damping and how it varies with TAS.
Using JSBSim to demonstrate ground effect during landing.
JSBSim has been used in a number of applications based on reinforcement learning.
Teams developed AI based fighter 'pilots/agents' with the winning team's agent beating a human fighter pilot 5-0 in mock air-combat. Youtube Video
MSc by Gordon Rennie.
This project aims at creating realistic open AI GYM environments using the open source Flight Dynamic Model JSBSim. Our work is initially based on Gordon Rennie's repository on which we made several modifications.
JSBSim is also used in academic and industry research, with over 600 references in Google Scholar. Here are a couple of interesting papers.
Flight Load Assessment for Light Aircraft Landing Trajectories in Windy Atmosphere and Near Wind Farms
This work focuses on the wake encounter problem occurring when a light, or very light, aircraft flies through or nearby a wind turbine wake. The dependency of the aircraft normal load factor on the distance from the turbine rotor in various flight and environmental conditions is quantified. For this research, a framework of software applications has been developed for generating and controlling a population of flight simulation scenarios in presence of assigned wind and turbulence fields. The JSBSim flight dynamics model makes use of several autopilot systems for simulating a realistic pilot behavior during navigation. The wind distribution, calculated with OpenFOAM, is a separate input for the dynamic model and is considered frozen during each flight simulation. The aircraft normal load factor during wake encounters is monitored at different distances from the rotor, aircraft speeds, rates of descent and crossing angles. Based on these figures, some preliminary guidelines and recommendations on safe encounter distances are provided for general aviation aircraft, with considerations on pilot comfort and flight safety. These are needed, for instance, when an accident risk assessment study is required for flight in proximity of aeolic parks. A link to the GitHub code repository is provided.
Evaluation of Formation Flights with Long Range Aircraft for Different Flight Conditions and Atmospheric Disturbances
The aerodynamic formation flight, which is also called air wake surfing for efficiency (AWSE), can lead to a high drag reduction at the trailing aircraft of more than ten percent resulting in a reduced fuel burn. Therefore, this operational strategy represents a promising means to reduce the greenhouse effect of aviation. The following study investigates the flight of two long haul commercial aircraft in an echelon formation in a stationary state, a flight dynamic simulation and finally at trajectory level. Thereby, the effect of different cruise altitudes and speeds, aircraft masses, lateral and vertical separations and different intensities of gusts and turbulence are evaluated. Based on the aircraft data set from the in house preliminary aircraft design tool MICADO a vortex lattice method calculates the induced loads in the trailing wake behind the leader. Subsequently, the results are used in the flight simulation program to analyze the flight behavior of the trailing aircraft in the formation under the influence of atmospheric disturbances. Finally, the results of the vortex lattice method and the flight simulation provide the necessary input data for the evaluation of the benefits achievable during the entire mission based on a detailed trajectory calculation. High altitudes and low Mach numbers during the formation flight lead to the highest drag reductions at the trailing aircraft. Movements of the trailing aircraft away from the optimum location in the vortex of the leading aircraft and additional detours lead to reduced fuel savings of ten percent or less.
The offline development, validation and verification of the software for an airborne strapdown inertial navigation system (SINS) requires that the simulation of the inertial measurement unit (IMU) be based on representative flight dynamics data. This paper presents our efforts to combine an existing, popular and open-source flight dynamics model (FDM) with an IMU simulator we developed for the purpose of such work. We will show how it is possible to use the FDM and IMU simulator to develop, evaluate and improve SINS software. This development also allows for validation and verification of the end software product.
Characterization of Derived Angle-of-Attack and Sideslip Angle Algorithms Using Monte Carlo and Piloted Simulation
Proper exploitation of derived angle-of-attack and sideslip angle from low cost Attitude Heading Reference System sensors found in General Aviation aircraft is a candidate solution for improving flight safety. This approach uses equations for angle-of-attack and sideslip angle that are solved onboard in real-time, using state data and information provided by Attitude Heading Reference System sensors. These equations are a function of vehicle parameters and stability and control derivatives, in addition to the angular displacement and angular rate sensor outputs. This work evaluates the feasibility of derived angle-ofattack and sideslip angle for General Aviation for use cases of pilot displays, envelope protection, and fly-by-wire flight control systems. The aircraft considered are Part 23 aircraft such as Cessna 172 and light jets. Standard angle-of-attack and sideslip angle equations are implemented in a nonlinear six degree-of-freedom simulation model of the Cessna 172, and the Attitude Heading Reference System is modeled on a generic type found in General Aviation aircraft. The Cessna 172 simulator is flown by evaluation pilots for the purpose of recording pilot input commands during the approach and landing flight phase. The pilot inputs are then used in a non real-time Monte Carlo flight simulation of the Cessna 172 that evaluates the effect of modeling uncertainties and sensor noise on derived angle-of-attack and sideslip angle. Monte Carlo cases of 10,000 and 100,000 runs are evaluated. Results presented in the paper show the allowable bounds on aircraft model parameter uncertainties such as stability and control derivatives, in addition to sensor noise levels, that produce usable derived angle-of-attack and sideslip angle values from low cost Attitude Heading Reference System sensors. It is expected that these results will be used for recommended minimum performance standards for the algorithm and Attitude Heading Reference System devices, in addition to the criteria for each use case when using Attitude Heading Reference Systems that can be codified into a standard or a circular.