A simulation based study of subliminal control for air traffic management

doi:10.1016/j.trc.2010.03.002

Transportation Research Part C: Emerging Technologies

Volume 18, Issue 6, December 2010, Pages 963-974

https://doi.org/10.1016/j.trc.2010.03.002 Get rights and content

Abstract

Under the so called Subliminal Control concept, an automated system, commanding minor speed adjustments imperceptible by the Air Traffic Controller (ATC), tries to keep the Air Traffic Controller’s risk perception low, emulating a “lucky traffic”. In this paper we outline such a concept and investigate several implementation considerations of subliminal control. A proposed subliminal controller is tested against several encounter geometries for level flights in simulations using a stochastic environment that comprises wind forecast uncertainties. The results demonstrate that subliminal control has the potential to reduce the workload of the ATC in several cases.

Introduction

The current Air Traffic Management (ATM) system is to a large extent based on a rigidly structured airspace and a mostly human-operated system architecture, see for example Tomlin, 1998, Nolan, 1998. For the separation assurance between aircraft, Air Traffic Controllers (ATC) have to make decisions in this highly uncertain and complex environment. To do so, they have to estimate the future positions of aircraft and intervene whenever they perceive a high risk of loss of separation. It is obvious that the traffic increase, predicted by European Commission, 2001, Eurocontrol Air Traffic Statistics and Forecast Service, 2004, demands an increase in the number of aircraft per sector. This will result in more stress on the ATC. To alleviate some of this workload, several alternative concepts of operation have been proposed, mostly centered around Conflict Detection and Resolution (CD&R) algorithms. For a thorough overview and classification of the literature on CD&R methods, the reader is referred to Kuchar and Yang (2000).

Subliminal control is one such concept, proposed in Villiers (2004). The premise is that minor speed adjustments, commanded by an automated system running in parallel with the ATC and using a dedicated ground-air datalink, can convert a potentially conflicting situation into a “lucky traffic” for the ATC, in the sense that most trajectories turn out to ensure safe separation at an early stage, thus reducing his monitoring workload. The speed adjustments have to be as small as possible in order to remain imperceptible by the ATC. In this approach the human is kept in the loop, and automation is introduced in a user-friendly way which reduces uncertainty without changing the work environment. Let us emphasize that the only effect of subliminal control from the ATC point of view is that traffic “behaves” better. From a control theory perspective, subliminal control can be viewed as an “inner-loop” controller, providing small scale adjustments in the system, improving the performance of the “outer-loop” of the system, the ATC, who performs coarser adjustments in the system. The hope is that well-behaved traffic can have higher density without straining the ATC capacity. The present paper is part of an on-going effort to investigate the potential of this approach.

The subliminal control concept was further developed under the project “En Route Air Traffic Soft Management Ultimate System” (ERASMUS) funded by European Commission (2006). As part of this effort, Crück and Lygeros (2007a) presented a mathematical framework for subliminal control. It relies on a model of the way ATC perceive the risk of loss of separation between aircraft as a function of the current traffic situation. The subliminal control problem is then modeled as a hybrid dynamical game in which the control has to minimize a cost representing the risk perceived by the ATC despite the uncertainty of trajectory prediction. The model of ATC’s risk perception is discussed in more details in Crück and Lygeros (2007b). Building on this work, the present paper develops a unified theoretical framework for quantitatively describing subliminal control concepts, so that their performance can be evaluated against different traffic scenarios. The evaluation is performed using Monte Carlo simulations running on the air traffic simulation platform presented in Lymperopoulos et al. (2007).

We present early results of our evaluations in two different sets of ATM scenarios: the current situation and the one envisioned by the SESAR initiative (see Ky and Miaillier, 2006). We consider two main differences between the current and future ATM system. The first is that in the SESAR framework, aircraft trajectories are highly constrained in time (4D trajectories), which reduces the margin of manoeuvre for speed variations for subliminal control. The second difference is that in the current framework, subliminal control speed variations have to be approved by the pilots onboard the aircraft, which yields a stochastic delay in execution; we assume that this delay is reduced in the SESAR framework thanks to better air-ground integration.

The rest of the paper is organized as follows: Section 2 introduces and discusses the subliminal control concept and our modeling framework. Section 3 describes in more details the modeling of ATC risk perception, a crucial parameter for the implementation and evaluation of subliminal control concepts. Section 4 presents the other aspects of the model we have used for the simulation results and for the concept validation. Section 5 briefly describes the simulation setup and presents the results of this study. Finally, Section 6 states the conclusions of this work.

Section snippets

Subliminal control concept

The main idea of subliminal control is to turn the uncertainty of the ATC about traffic evolution into an advantage. It has been shown that small adjustments of speeds commanded early enough can prevent a large percentage of conflicts (Alliot et al., 2001). Subliminal control considers speed adjustments small enough to be within the uncertainty margin of the ATC and hence, in principle, imperceptible. Results from human in the loop experiments of the European project ERASMUS (European

Risk model

The risk perception model we use is along the lines of the one proposed in Crück and Lygeros (2007b). The risk function for a given traffic situation is defined as: $R_{ATC, n} : X_{1} \times X_{2} \times \dots \times X_{n} \to [0, 7],$ where $X_{i}$ denotes the state of aircraft i and n is the total number of aircraft in the traffic situation. The range of values [0, 7] has been set in reference to the experimental setting in Averty et al. (2006) which is, to the best of our knowledge, the only available study attempting to quantify ATC risk

Aircraft/flight environment model

The risk perception model of Section 1 and its implications for subliminal control decision making were evaluated by simulations. The model developed in Lymperopoulos et al. (2007) was used for this purpose. This model allows one to capture multiple flights taking place at the same time. Each flight has an associated flight plan, aircraft dynamics and a Flight Management System (FMS). The evolution of flights is affected by the wind speed, which is uncertain, making the model stochastic.

Simulation results

In this section we examine several two aircraft encounters to determine the minimum time $T_{horizon}$ for a subliminal control scheme to keep the risk perception of the ATC low for each encounter. We consider two aircraft flying level at the same altitude, in straight lines, at constant airspeeds (see Fig. 7). In the absence of uncertainty, the minimum distance that the two aircraft approach each other is denoted $δ_{\min}$ and the time this event occurs $t_{\min}$ (time to minimum separation).

We construct

Conclusions and future work

We have investigated the potential of the use of subliminal control to alleviate ATC workload involved in monitoring some potentially dangerous encounters. For simple two-aircraft situations, the results clearly indicate that, subliminal control can reduce the workload of the ATCs monitoring situations. The efficiency of subliminal control depends of course on the nominal minimal separation between the aircraft and on the time available to solve the situation before it is perceived as a medium

Acknowledgment

This research is supported by the European Commission under the project ERASMUS, FP6-TREN-518276.

References (21)

Alliot, J.-M., Durand, N., Granger, G., 2001. A statistical analysis of the influence of vertical and ground speed...
Averty, P., Guittet, K., Lezaud, P., July 2006. Perception of Risks of Conflict by Air Traffic Controllers: The CREED...
G. Chaloulos et al.
Effect of wind correlation on aircraft conflict probability
AIAA Journal of Guidance, Control and Dynamics
(2007)
Cole, R., Richard, C., Kim, S., Bailey, D., July 1998. An assessment of the 60km rapid update cycle (RUC) with near...
Crück, E., Lygeros, J., August 2007a. A mathematical framework for subliminal air traffic control. In: AIAA Guidance,...
Crück, E., Lygeros, J., July 2007b. Hybrid modeling for the evaluation of risk perception by air-traffic controllers....
Eurocontrol Air Traffic Statistics and Forecast Service, February 2004. Forecast of Annual Number of IFR Flights...
Eurocontrol Experimental Centre, 2004. User Manual for the Base of Aircraft Data (BADA)....
European Commission, 2001. European Transport Policy for 2010: Time to Decide....
European Commission, 2006. ERASMUS – En Route Air Traffic Soft Management Ultimate System....

There are more references available in the full text version of this article.

Cited by (30)

Applications of stochastic modeling in air traffic management: Methods, challenges and opportunities for solving air traffic problems under uncertainty
2021, European Journal of Operational Research
Citation Excerpt :
In addition, the modeling of wind conditions is often underpinned by Gaussian processes, although it is important to take correlation structure into account in such models. A typical approach, as used in Chaloulos, Cruck, and Lygeros (2010), is to model wind velocity as the sum of a deterministic (nominal) component and a stochastic component, where the stochastic component has the form of a Gaussian random field. This type of model can ensure wind correlation in time and space; in other words, the wind conditions experienced by an aircraft at a particular time are correlated with those of other aircraft at the same time, and also with the wind conditions at earlier times.
In this paper we provide a wide-ranging review of the literature on stochastic modeling applications within aviation, with a particular focus on problems involving demand and capacity management and the mitigation of air traffic congestion. From an operations research perspective, the main techniques of interest include analytical queueing theory, stochastic optimal control, robust optimization and stochastic integer programming. Applications of these techniques include the prediction of operational delays at airports, pre-tactical control of aircraft departure times, dynamic control and allocation of scarce airport resources and various others. We provide a critical review of recent developments in the literature and identify promising research opportunities for stochastic modelers within air traffic management.
Learning aircraft operational factors to improve aircraft climb prediction: A large scale multi-airport study
2018, Transportation Research Part C: Emerging Technologies
Citation Excerpt :
As an example, in Prats et al. (2010) an iterative quasi-Newton method is used to find trajectories for departing aircraft, minimizing the noise annoyance. Another example is Chaloulos et al. (2010) where Monte Carlo simulations are used to estimate the risk of conflict between trajectories, in a stochastic environment. Some of the automated tools currently being developed for ATC/ATM can detect and solve conflicts between trajectories, using Genetic Algorithms (Durand et al., 19961), or Differential Evolution or Particle Swarm Optimization (Vanaret et al., 2012).
Ground-based aircraft trajectory prediction is a major concern in air traffic control and management. A safe and efficient prediction is a prerequisite to the implementation of new automated tools.
In current operations, trajectory prediction is computed using a physical model. It models the forces acting on the aircraft to predict the successive points of the future trajectory. Using such a model requires knowledge of the aircraft state (mass) and aircraft intent (thrust law, speed intent). Most of this information is not available to ground-based systems.
This paper focuses on the climb phase. We improve the trajectory prediction accuracy by predicting some of the unknown point-mass model parameters. These unknown parameters are the mass and the speed intent. This study relies on ADS-B data coming from The OpenSky Network. It contains the climbing segments of the year 2017 detected by this sensor network. The 11 most frequent aircraft types are studied. The obtained data set contains millions of climbing segments from all over the world. The climbing segments are not filtered according to their altitude. Predictive models returning the missing parameters are learned from this data set, using a Machine Learning method. The trained models are tested on the two last months of the year and compared with a baseline method (BADA used with the mean parameters computed on the first ten months). Compared with this baseline, the Machine Learning approach reduce the RMSE on the altitude by 48% on average on a 10 min horizon prediction. The RMSE on the speed is reduced by 25% on average. The trajectory prediction is also improved for small climbing segments. Using only information available before the considered aircraft take-off, the Machine Learning method can predict the unknown parameters, reducing the RMSE on the altitude by 25% on average.
The data set and the Machine Learning code are publicly available.
Adaptive prediction of flight time uncertainty for ground-based 4D trajectory management
2018, Transportation Research Part C: Emerging Technologies
Citation Excerpt :
It was mentioned that the position uncertainty also depends on the meteorological conditions (Bakker et al., 2000), and a “dynamic spacing” concept was presented in which position uncertainty prediction based on meteorological conditions could improve both the safety and efficiency of air traffic management. Despite the existence of this concept, the static modeling of trajectory uncertainty has been widely applied in studies in the air traffic management field (Chaloulos and Crück, 2010; Lymperopoulos and Lygeros, 2010; Margellos and Lygeros, 2013; Weitz, 2013). Regarding trajectory uncertainty prediction based on meteorological conditions, the modeling of wind uncertainty correlation between aircraft pairs (Hu et al., 2005; Chaloulos and Lygeros, 2007) and its application for conflict resolution (Matsuno et al., 2015) has been discussed.
An adaptive prediction model of level flight time uncertainty is derived as a function of flight and meteorological conditions, and its effectiveness for ground-based 4D trajectory management is discussed. Flight time uncertainty inevitably increases because of fluctuations in meteorological conditions, even though the Mach number, flight altitude and direction are controlled constant. Actual flight data collected using the secondary surveillance radar Mode S and numerical weather forecasts are processed to obtain a large collection of flight time error and flight and meteorological conditions. Through the law of uncertainty propagation, an adaptive prediction model of flight time uncertainty is derived as a function of the Mach number, flight distance, wind, and temperature. The coefficients of the adaptive prediction model is determined through cluster analysis and linear regression analysis. It is clearly demonstrated that the proposed adaptive prediction model can estimate the flight time uncertainty without underestimation or overestimation, even under moderate or severe weather conditions. The proposed adaptive prediction is able to improve both safety and efficiency of 4D trajectory management simultaneously.
A dynamic air traffic model for analyzing relationship patterns of traffic flow parameters in terminal airspace
2016, Aerospace Science and Technology
Citation Excerpt :
From separation maintaining point of view, mostly centered around are the conflict detection and resolution models. A thorough overview and classification of these models was presented by Kuchar and Yang [23], while more recently, the reader is referred to [5,47,40]. With the improvement in simulation techniques, air traffic models were proposed and modified with respect to specific scenarios and constraints.
This paper provides a special view on air traffic flow parameters and presents certain influences from airspace to traffic flow characteristics. Based on statistics on measurement radar data within terminal airspace, we obtain time distributions and interrelationships of the basic parameters, and accordingly divide traffic flow states into 3 phases: free flow, mild-controlled and strong-controlled flow. Through analyzing characteristics shown in different phase states, as well as aircraft performing in real airspace operations, we establish dynamic models to describe potential behaviors including following, holding and maneuvering, while programing them into simulation tools. Relationship patterns are validated by comparison analysis on the similar simulation results against measurement data. Taking air route adjustment as simulation example, we conduct experiments on airspace elements that may have effects on traffic flow, with further discussing on possible reasons.
Operating cost sensitivity to required time of arrival commands to ensure separation in optimal aircraft 4D trajectories
2015, Transportation Research Part C: Emerging Technologies
Citation Excerpt :
Another example is the project Environmentally Responsible Air Transport (ERAT), which developed methods and systems to reduce environmental impact using continuous arrival and departure procedures inside high density TMAs in Europe (ERAT Consortium, 2012). Alternatively, Chaloulos et al. (2010) propose imperceptible speed adjustments to minimise the risk of potential conflicts and lessen the workload of ATCo, impacting positively into airspace capacity. The current paradigm for air traffic management involves strict air traffic control (ATC) procedures and lacks advanced automation.
Trajectory optimisation has shown good potential to reduce environmental impact in aviation. However, a recurring problem is the loss in airspace capacity that fuel optimal procedures pose, usually overcome with speed, altitude or heading advisories that lead to more costly trajectories. This paper aims at the quantification in terms of fuel and time consumption of implementing suboptimal trajectories in a 4D trajectory context that use required times of arrival at specific navigation fixes. A case study is presented by simulating conflicting Airbus A320 departures from two major airports in Catalonia. It is shown how requiring an aircraft to arrive at a waypoint early or late leads to increased fuel burn. In addition, the efficiency of such methods to resolve air traffic conflicts is studied in terms of both fuel burn and resulting aircraft separations. Finally, various scenarios are studied reflecting various airline preferences with regards to cost and fuel burn, as well as different route and conflict geometries for a broader scope of study.
Stochastic optimal control for aircraft conflict resolution under wind uncertainty
2015, Aerospace Science and Technology
In this paper, a stochastic optimal control method is developed for determining three-dimensional conflict-free aircraft trajectories under wind uncertainty. First, a spatially correlated wind model is used to describe the wind uncertainty, and a probabilistic conflict detection algorithm using the generalized polynomial chaos method is proposed. The generalized polynomial chaos algorithm can quantify uncertainties in complex nonlinear dynamical systems with high computational efficiency. In addition, a numerical algorithm that incorporates the generalized polynomial chaos method into the pseudospectral method is proposed to solve the conflict resolution problem as the stochastic optimal control problem. The stochastic optimal control method is combined with the proposed conflict detection algorithm to solve the conflict resolution problem under the wind uncertainty. Through illustrative three-dimensional aircraft conflict detection and resolution examples with multiple heterogeneous aircraft, the performance and effectiveness of the proposed conflict detection and resolution algorithms are evaluated and demonstrated.

View all citing articles on Scopus

View full text