Search

Accurate air traffic flow prediction assists controllers formulate control strategies in advance and alleviate air traffic congestion, which is important to flight safety. While existing works have made significant efforts in exploring the high dynamics and heterogeneous interactions of historical air traffic flow, two key challenges still remain. (1) The transfer patterns of air traffic are intricate, subject to numerous constraints and limitations such as controllers, flight regulations, and other regulatory factors. Relying solely on mining historical traffic evolution patterns makes it difficult to accurately predict the constrained air traffic flow. (2) Weather conditions exert a substantial influence on air traffic, making it exceptionally difficult to simulate the impact of external factors (such as thunderstorms) on the evolution of air traffic flow patterns. To address these two challenges, we propose a Spatiotemporal Knowledge Distillation Network (ST-KDN) for air traffic flow prediction. Firstly, recognizing the inherent future insights embedded within flight plans, we develop a "teacher-student" distillation model. This model leverages the prior knowledge of upstream-downstream migration patterns and future air traffic trends inherent in flight plans. Subsequently, to model the influence of external factors and predict air traffic flow disturbed by thunderstorm weather, we propose a student network based on the "parallel-fusion" structure. Finally, employing a feature-based knowledge distillation approach to integrate prior knowledge from flight plans and extract meteorological features, our method can accurately capture complex and constrained spatiotemporal dependencies in air traffic and explicitly model the impact of weather on air traffic flow. Experimental results on real-world flight data demonstrate that our method can achieve better prediction performance than other state-of-the-art comparison methods, and the advantages of the proposed method are particularly prominent in modeling the complicated transfer pattern of air traffic and inferring nonrecurrent flow patterns. [ABSTRACT FROM AUTHOR]

Flight sequence is essential for aviation's secure and orderly operation. Accurate Flight Sequence Prediction (FSP) gives air traffic controllers (ATCs) an overview of the overall air traffic situation during the pre-tactical phase, allowing them to modify flight schedules when airspace capacity is reduced. The internal structure of flight sequences is compact, and then the effective extraction of the spatiotemporal dependence features of flight sequences determines the accuracy of the prediction. The flight scheduling approach treats the FSP as an optimization problem with the aircraft already known, but when considering more airlines, the problem becomes complicated to solve. Causal inference methods for solving FSP have struggled to extract implicit interaction features between flight sequences. For the above bottlenecks, we construct flight sequences from the perspective of sequence contents and transform the FSP problem into a session-based recommendation task by proposing a novel Adaptive Airport Awareness Graph Neural Network (AAGNN). Firstly, we lead in all the sequences to form an airport network, and then we augment the graph topology of the flight sequences with network structure information. Secondly, we propose an airport-aware adaptive representation graph model to obtain the implicit interaction features between flight sequences at the same airport. Finally, we develop a hybrid embedded session representation incorporating recent and overall flight sequence preferences for session recommendation. The experimental results on two real-world datasets of flight sequences in China demonstrate that the proposed AAGNN approach outperforms the numerous baseline approaches. The case study demonstrates the approach's generalizability to various specific operational scenarios. • Flight sequence prediction transformed into a session-based recommendation task. • Airport network features are used to model the interactions among flight sequences. • The adaptive method guarantees the sequence update as network properties change. • The proposed approach outperforms on two real-world datasets in China. [ABSTRACT FROM AUTHOR]

In this paper, we study the integrated production-maintenance optimization problem of a deteriorating machine where random yield and maintenance delay are considered. The system deterioration is described by a discrete-time Markov chain. The production system faces a constant demand and a random yield which is proportional to the input quantity. A corrective maintenance needs to be scheduled if the machine enters into its failure state. Otherwise, the decision-maker can choose to schedule a preventive maintenance or decide how much to produce. The maintenance delay is considered in this work, where the maintenance crew arrival time is assumed to be non-negligible. We formulate the problem into a Markov decision process framework where the total discounted production-maintenance costs in the infinite horizon is minimized. Some structural properties of the optimal policy with respect to the machine condition, the inventory level are presented under mild conditions. A numerical example is given to present the model applicability. We also compare the model with the sequential approach to illustrate the characteristics and advantages of the proposed model. It may provide some managerial insight to the decision-maker when developing production schedules. • An integrated production and maintenance planning of a degrading system is studied. • The impact of maintenance delay and random yield are examined. • Some structural properties of the joint policy are derived. [ABSTRACT FROM AUTHOR]

In this paper, we study the integrated production–maintenance optimization problem of a multi-component deteriorating machine. The deterioration of each component is described by a discrete-time Markov chain with finite state space. To meet a constant demand, production planning is scheduled based on the system deterioration and the inventory level. We consider the mutual dependence between the production and the system deterioration. For each component, its deterioration transition probabilities depends on its own characteristics and its production quantity. Production yield is stochastically decreasing with the system deterioration. Maintenance is scheduled immediately if at least one component failure occurs. Otherwise, the decision-maker need to determine whether to schedule a preventive maintenance or to continue producing. We formulate the problem into a Markov decision process framework. The total discounted costs including the production costs, maintenance costs and holding/backlogging costs in the infinite horizon is obtained. Some structural properties of the optimal policy with respect to the machine condition, the inventory level are presented under mild conditions. The proposed model is further examined by a numerical example, where the properties of the production and maintenance planning and the corresponding cost are investigated. It can provide theoretical reference in managing the production and maintenance problems in multiple production lines. • We study an integrated production–maintenance optimization problem of a parallel production system with random yield. • We model and examine the dependence between production quantity and system deterioration. • We investigate structural properties of the joint policy in the finite-horizon and infinite-horizon respectively. • The impact of random yield, production–deterioration dependence on the optimal production and maintenance planning are illustrated. [ABSTRACT FROM AUTHOR]

This paper presents a novel hybridized indirect and direct encoding (HybrID) genetic algorithm for solving air traffic network flow optimization problems. A heuristic, which uses the Dijkstra algorithm for generating different types of shortest paths on a graph while controlling the weights on each arc, is proposed for selecting optimal flight routes based on current air traffic. A novel HybrID chromosome representation is employed along with the proposed heuristic and a genetic algorithm for optimization. Experiments on synthetic problems and real data of the Chinese airspace show the proposed method outperforms the direct encoding method on efficiency and efficacy metrics. [ABSTRACT FROM AUTHOR]

This article concentrates on the challenging adaptive tracking problem of multi‐input and multi‐output (MIMO) nonlinear systems with unknown nonlinear dynamics, for which a novel optimization‐based data‐driven adaptive control (ODDAC) equipped with an extended dynamic linearization method is developed. Through considering MIMO nonlinear systems in the fully‐actuated case and over‐actuated case separately, the ODDAC is proposed consisting of a parameter updating algorithm and an adaptive control law. A new design of parameter updating algorithm is presented such that the estimation can be guaranteed to be bounded by a strict contraction process. By leveraging properties of the nonnegative matrix, a grouping‐based contraction mapping (GCM) analysis method is proposed for the convergence of tracking error. Notably, the GCM does not require the contraction mapping condition to hold at all time instants. The proposed ODDAC is data‐based, avoiding reliance on model information, and its validity is verified through simulations. [ABSTRACT FROM AUTHOR]

In this article, the condition-based maintenance optimization of a K-out-of-N deteriorating system considering failure dependence is discussed. The degradation of each component is modelled by a pure jump Lévy process. Whenever one component fails, it can either induce instantaneous failures or lead to the increment of degradation levels of other components. Thus, this model has the flexibility to describe the phenomena of instantaneous failures of multiple components, which is known as the common cause failure. It can also model the accumulative, gradual propagation effect of the component failure to the system. A periodic inspection policy is considered to reveal the real state of the system, upon which, possible maintenance actions can be carried out according to the observations. The inspection and maintenance problem is formulated as a Markov decision process and the value iteration algorithm is employed to solve the problem. The proposed policy is assessed by the total expected discounted cost in the long-run horizon. Under mild conditions, some structural properties of the optimal maintenance policies are obtained. A numerical example is given to illustrate the applicability of the proposed model. It can provide theoretical reference for the decision-maker when developing maintenance policies. [ABSTRACT FROM AUTHOR]

In this paper, a condition-based maintenance of a two-component system under imperfect inspection is studied: component 1 is repairable and only two states (working and failed) are observable. Component 2 degrades with time. The degradation of component 2 is modelled by a Wiener process with different drift and volatility parameters depending on the working state of component 1. Component 2 fails if its degradation level exceeds a pre-defined threshold. A dual periodical inspection policy is proposed to monitor the system condition: component 1 is imperfectly inspected and the whole system is perfectly inspected with a longer period. Upon inspection, component 1 is replaced if a failure is detected. The whole system is renewed if the degradation of component 2 exceeds its preventive maintenance threshold. The problem is very complex due to the failure dependence between the two components and the partial revealed information within the system inspection period. Component 1 is impacted by the imperfect inspection. Component 2 may experience multiple changes of its degradation regime while only imperfect, partial change information is available. In this study, we intend to address these problems. The expression of the maintenance cost rate in the long-run is given to evaluate the maintenance policy. An algorithm is proposed to achieve the maintenance optimization. Numerical examples show the applicability of the model. It can provide reference to the decision-maker for optimizing the maintenance budget. • A condition-based maintenance of a two-unit system with failure dependence is studied. • A dual inspection policy considering the imperfect inspection is proposed. • A joint optimization of the inspection and the maintenance threshold is presented. • An algorithm is presented to recursively solve the optimization problem. [ABSTRACT FROM AUTHOR]

In this article, a new data‐based iterative learning control (ILC) algorithm is proposed via Gaussian process regression (GPR) to accomplish the trajectory tracking objective of aircraft subject to completely unknown dynamics and strong nonlinearities. The nonlinear system input–output relationship of the unknown aircraft is formulated through GPR by leveraging historical data, based on which an optimal ILC framework is established. The monotonic convergence analysis of the GPR‐based ILC is explored such that high‐precision tracking tasks can be accomplished without prior model knowledge. Simulation tests are further conducted on a commercial aircraft performing a continuous climb operation to illustrate the effectiveness of the GPR‐based ILC approach. [ABSTRACT FROM AUTHOR]

Robust gate assignment problem (RGAP) aims to generate a reliable schedule that reduces the occurrence of potential gate conflicts under stochastic flight delays. Although allocating buffer time in the gate schedules is widely adopted to absorb the delay disturbances, the relation between the schedule robustness and the aircraft delay heterogeneity has not yet been well explored. To allocate the buffer optimally, we analyse the heterogeneity of gate conflict and propose two new robustness measurements to quantify the number of potential conflicts from the expectation perspective. Then the RGAP is modelled as a set-covering problem to simultaneously optimise the expected conflict cost and basic operating cost. To solve the model efficiently, we develop a column generation-based heuristic (CDBH) to find high-quality solutions. Computational studies based on Shenzhen Baoan International Airport reveal that the proposed measurements considering the delay heterogeneity can significantly improve the robustness of gate schedules. [ABSTRACT FROM AUTHOR]

This paper contributes to solving a class of so-called ‘transcale average consensus’ problems for directed networks of vehicles with single-integrator dynamics. It generalises the idea of average consensus by taking into account the measurement scale of every vehicle's information state. Using the nearest neighbour-interaction rules, distributed algorithms are presented which are shown with the ability to achieve the transcale average consensus for directed networks associated with balanced graphs. Such consensus results are applicable to networks in the presence of both fixed topology and switching topologies, which are also illustrated via simulation tests. Moreover, improved algorithms and results of consensus are given for directed networks associated with unbalanced graphs. [ABSTRACT FROM PUBLISHER]

Air traffic complexity, which measures the disorder of air traffic distribution, has become the critical indicator to reflect air traffic controller workload in air traffic management (ATM) system. However, it is hard to assess the system accurately because there are too many correlated factors, which make the air traffic complexity nonlinear. This paper presents an air traffic complexity evaluation model with integrated classification using computational intelligence (ATCEM). To avoid redundant factors, critical factors contributing to complexity are analyzed and selected from numerous factors in the ATCEM. Subsequently, to construct the mapping relationship between selected factors and air traffic complexity, an integrated classifier is built in ATCEM. With efficient training and learning based on aviation domain knowledge, the integrated classifier can effectively and stably reflect the mapping relationship between selected factors and the category of air traffic complexity to ensure the precision of the evaluation. Empirical studies using real data of the southwest airspace of China show that the ATCEM outperforms a number of state-of-the-art models. Moreover, using the critical complexity factors selected in ATCEM, the air traffic complexity could be effectively estimated. Copyright © 2015 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

The robustness of realistic infrastructure systems is a significant and long-term research question, which has gotten some constructive outcomes borrowing the framework of complex networks. In this paper, we further explore the method about utilizing optimization algorithms to enhance robustness of interdependent networks. Firstly, we propose a novel particle swarm optimization algorithm with an information feedback mechanism to achieve better performance on complex problems. Secondly, we analyze the solutions from initial loads and types of failed nodes resulted by our algorithm. Interestingly, relative to traditional setup, our algorithm indeed produces a robust resource allocation pattern for interdependent networks, where middle-degree nodes are assigned more tolerance. [ABSTRACT FROM AUTHOR]

Conflict avoidance plays a crucial role in guaranteeing the safety and efficiency of the air traffic management system. Recently, the strategic conflict avoidance (SCA) problem has attracted more and more attention. Taking into consideration the large-scale flight planning in a global view, SCA can be formulated as a large-scale combinatorial optimisation problem with complex constraints and tight couplings between variables, which is difficult to solve. In this paper, an SCA approach based on the cooperative coevolution algorithm combined with a new decomposition strategy is proposed to prevent the premature convergence and improve the search capability. The flights are divided into several groups using the new grouping strategy, referred to as the dynamic grouping strategy, which takes full advantage of the prior knowledge of the problem to better deal with the tight couplings among flights through maximising the chance of putting flights with conflicts in the same group, compared with existing grouping strategies. Then, a tuned genetic algorithm (GA) is applied to different groups simultaneously to resolve conflicts. Finally, the high-quality solutions are obtained through cooperation between different groups based on cooperative coevolution. Simulation results using real flight data from the China air route network and daily flight plans demonstrate that the proposed algorithm can reduce the number of conflicts and the average delay effectively, outperforming existing approaches including GAs, the memetic algorithm, and the cooperative coevolution algorithms with different well-known grouping strategies. [ABSTRACT FROM AUTHOR]