Félix Mora-Camino

A soft dynamic programming approach for on-line aircraft 4D-trajectory optimization

During the last decade, Flight Management Systems (FMS) have been introduced on board modern aircraft to reduce operational costs, and flight profiles have been optimized in order to minimize fuel and time-related costs. Today, the integration of next generation FMS in the future air traffic management environment requires the development of new capabilities and among them, on-line real time 4D-trajectory generation has become mandatory. This paper presents a method based on dynamic programming (DP) to generate optimal 4D-trajectories in the presence of multiple time constraints. Soft computing techniques are introduced to improve the DP oriented optimization process by on-line selection of the search space and by cutting down computation time in repetitive performance evaluations.

Decentralized predictive controller for delivery canals

Presents the design of a predictive controller for decentralized control of an irrigation delivery canal. The overall canal is decomposed into single subsystems (pools) separated by a control structure (gate). The subsystems are inter-dependent because flow through the structures depends on both upstream and downstream water level. In the paper, predictive control theory is used to formulate a set of local controllers for the automatic operation of gates of each pool. The control scheme presented takes into account explicitly the interactions among subsystems. The objective of each controller is to maintain the downstream end water level of each pool at a constant target value under external perturbation. To evaluate the controller performance in realistic conditions, it is tested on an open-canal flow nonlinear model.

An intelligent decision support system for irrigation system management

Considers the design of a decision support system for the short term water resource management of an irrigation system. The operations of similar systems are often impaired by different stochastic events like device failure, heavy rains or dry periods and new long term goals. To be effective, such a decision support system which is based on knowledge techniques (state identification) and adaptive optimization (short term plans), requires the development of an information system based on water resource demand and supply. This information system gathers data from different fields (hydrology, meteorology and agriculture) so that accurate predictions about available reserves and demand levels can be performed. So, this communication presents the structure of the decision support system and focuses on tactical management information needs. The case study considered deals with a three-reach irrigation system.

Long-term fuzzy management of water resource systems

In the present context of water resource scarcity, a complete approach for long-term storage/transfer/distribution system management is proposed. The main management objective of such a kind of system is to manage reserves and releases so as to minimize the deficit between supply and demand by taking into account demand and contribution prediction. Thus, a sliding horizon approach and especially an adaptation procedure of weighting parameters of the minimization criterion based on fuzzy logic is implemented. This criterion adaptation concept seems entirely judicious, in view of the difficulty to define weighting parameters of such optimization problems and their influence on the relevant solution obtained. Fuzzy logic is shown to be very adequate when it comes to apprehend finely the stakes in presence in the long-term management. The long-term management problem is solved with improved heuristics using linear programming and dynamic programming in order to reduce the effects of space discretization which is so restrictive in this context. The suggested approach is applied to a case study which permits to highlight its relative simplicity of implementation.

Flexible management of water resource systems

This communication consider the problem of the management on the short-term of water resource systems (WRS). This problem is coped through an adaptive generation of a recurrent optimization problem which is solved by a specialized primal linear programming technique. A fuzzy approach is used in each instance of the recurrent optimization problem to express accurately the demand constraints and the terminal level stock constraints. The proposed approach, applied to a WRS of medium complexity, shows that it can be adapted to the management of particular situations like flood management, management in presence of failed devices,...

Decentralized adaptive predictive control of multireach irrigation canal

This paper presents the design of a predictive controller for decentralized control of an irrigation canal. The overall canal is decomposed into N subsystems (reaches) separated by a control structure (gate). The subsystems are interdependent because flow through the structures depends on both up- and downstream water levels. Here, predictive control theory is used to formulate a set of local controllers for the automatic operation of each reach gate. The control scheme presented explicitly takes into account the interactions among subsystems. The objective of each controller is to maintain the downstream end water level of each reach at a constant target value under external perturbation. To evaluate the controller performance in realistic conditions, it is tested on an open-canal flow non-linear model. The case study considered deals with eight canal reaches.

Automatic collision avoidance based on supervised predictive controllers

Abstract In general, adaptive predictive control techniques have been applied to systems presenting slow dynamics or large delays. In cases where fast maneuvers are considered, their parameters estimation function and their control computation function must be enhanced to cope with these transient situations. Here a supervision layer, based on fuzzy logic techniques, is introduced to realize the on-line adaptation of identification and control law parameters. This approach is applied to the case of aircraft collision avoidance maneuvers.

A maximum entropy approach to update airlines demand distributions

In order to design an efficient market-reactive revenue management system for airlines, it is necessary to have at hand a representative probability distribution of demand by period, by fare class and by order of arrival. For this, a dual geometric programming problem can be formulated according to the principle of maximum entropy. Making use of the corresponding primal form of the geometric program, the large-scale convex optimization problem is transformed into a non-constrained non-convex minimization problem. The solution strategy proposed consists of two main steps: first, solve the primal form of the geometric program and second, compute by geometric inversion the updated probability distributions. A numerical solution of the non-convex primal geometric program is obtained using an ad hoc designed genetic algorithm. Its performances had been evaluated under different simulation scenarios involving various fare classes, several forecasting periods and different demand profiles, showing satisfactory results.

Aircraft-on-Ground Lateral Control for Low Speed Maneuvers

Abstract The paper presents the study of a yaw rate control of the aircraft-on-ground, based on feedback linearizing. This technique aims at designing a non-linear controller that constrains the system outputs to follow a linear reference behavior. However, it assumes that the reference model remains valid for all the operating range. This assumption is not verified for very low and null speeds. For such speeds, even the control objective is not operationally acceptable. Therefore, this communication proposes adaptations of the controller to follow the initial requirements in all the speed range.

Control of Input Redundant Affine Systems: Application to Runway Axis Tracking at Landing

Among the specific methodologies for the control of nonlinear systems, Nonlinear Dynamic Inversion has gained considerable popularity over the last decade. A complete theory is now available for the design of feedback control laws which associates to a nonlinear system a dynamic linear input-output relationship easily inversible. Nevertheless, the possible saturation of the actuators has not been taken into account. In order to solve this limitation, a supervision strategy is proposed in the case of input redundant affine systems. In this paper, while a generic force and moment nonlinear inverse control laws is displayed, the viability and efficiency of the proposed supervision strategy is highlighted.