Search

We investigate a general solution based on reinforcement learning for the agile satellite scheduling problem. The core idea of this method is to determine a value function for evaluating the long-term benefit under a certain state by training from experiences, and then apply this value function to guide decisions in unknown situations. First, the process of agile satellite scheduling is modeled as a finite Markov decision process with continuous state space and discrete action space. Two subproblems of the agile Earth observation satellite scheduling problem, i.e., the sequencing problem and the timing problem are solved by the part of the agent and the environment in the model, respectively. A satisfactory solution of the timing problem can be quickly produced by a constructive heuristic algorithm. The objective function of this problem is to maximize the total reward of the entire scheduling process. Based on the above design, we demonstrate that Q-network has advantages in fitting the long-term benefit of such problems. After that, we train the Q-network by Q-learning. The experimental results show that the trained Q-network performs efficiently to cope with unknown data, and can generate high total profit in a short time. The method has good scalability and can be applied to different types of satellite scheduling problems by customizing only the constraints checking process and reward signals.

The scheduling of agile Earth observation satellites is to select a subset of candidate targets each associated with a profit during their visible time windows in order to maximize the collected profits, under some operational constraints. For each pair of two consecutive observations, a transition time is required to perform a rotating movement of the camera, depending on the start times of the two observations. This time-dependency significantly increases the complexity of the scheduling problem. To solve this problem efficiently, we model the time-dependent transition time and prove that it satisfies the first-in-first-out rule and the triangle inequalities rule. On this basis, we develop a novel hybrid heuristic, called "greedy randomized iterated local search'' (GRILS). A specific insert operator, including a fast feasibility check and an assignment procedure are specifically designed to address the operational constraints of the scheduling. Extensive experiments on the single satellite instances and multisatellite instances demonstrate that our algorithm outperforms the state-of-the-art algorithms with respect to solution quality and computation time. ispartof: IEEE Transactions on Systems, Man and Cybernetics: Systems vol:52 issue:3 pages:1614-1625 status: published

In recent years, research into self-organized swarm robotics has received much attention. Within this research, development of a general control model that functions under different conditions has been a hot topic. One significant approach is to use a control model with tunable parameters determined through simulation. In this article, we propose a rule-based swarm robotics control model with tunable parameters. To improve the performance, a modified non-dominated sorting genetic algorithm (NSGA II) is embedded into this control model. Multiple scenarios including convex obstacles, square obstacles, and tunnels are used to test the performances of the proposed control model. Comparison is made with three widely applied population-based multi-objective optimization algorithms. Experimental results show that our control model has good performance and is robust under different scenarios.

The effective scheduling of multiple agile earth observation satellites plays a key role in improving the efficiency of the agile satellite observation system. Due to the complex constraints and large solution space (increases exponentially with the problem size), it has been a big challenge to effectively solve the multiple agile satellites scheduling problem. In this study, an edge computing framework is proposed in order to solve this problem in a flexible manner. In this framework, a central node and several edge nodes are included. The central node corresponds to the operation center of the satellite observation system while each edge node corresponds to an agile satellite. The central node filters tasks for edge nodes according to the time of scheduling period and visible time windows. Tasks in each edge node are scheduled according to the order of scheduling periods by a constructive heuristic algorithm based on the density of residual tasks (HADRT). The efficiency of this algorithm is proved. The time and space complexity are analyzed in detail. The proposed method is compared with other three advanced methods, and the experimental results show that HADRT could achieve higher task completion rate and reward rate than other algorithms with acceptable computing time.

To study the influence of different factors on the fuel consumption of vehicles on the superhighway, and determine the weight of the influence of different factors and the relationship among the factors. Through the literature research and Delphi method, 16 factors affecting fuel consumption were determined, which were divided into four dimensions: driver, vehicle, road, and environment. Integrated Decision-Making and Trial Evaluation Laboratory (DEMATEL) and interpretive structural model (ISM) was used to establish a multi-level hierarchical structure model. The results show that the hierarchical structure model can be divided into six layers. The first layer is the driving technology and driving speed, which are the direct factors affecting the fuel consumption of vehicles; the sixth layer is the engine performance and topographic characteristics, which is the deep factor affecting the fuel consumption of vehicles. Besides, by calculating the cause degree and the center degree for analysis, it is found that there are 10 casual factors, 6 results factors, and 8 key influencing factors. The research results can provide a reference for the analysis of vehicle fuel consumption on the superhighway and provide a reference for the construction of the superhighway with a superior economy.

In this paper, we study a new representation-learning task, which we termed as disassembling object representations. Given an image featuring multiple objects, the goal of disassembling is to acquire a latent representation, of which each part corresponds to one category of objects. Disassembling thus finds its application in a wide domain such as image editing and few- or zero-shot learning, as it enables category-specific modularity in the learned representations. To this end, we propose an unsupervised approach to achieving disassembling, named Unsupervised Disassembling Object Representation (UDOR). UDOR follows a double auto-encoder architecture, in which a fuzzy classification and an object-removing operation are imposed. The fuzzy classification constrains each part of the latent representation to encode features of up to one object category, while the object-removing, combined with a generative adversarial network, enforces the modularity of the representations and integrity of the reconstructed image. Furthermore, we devise two metrics to respectively measure the modularity of disassembled representations and the visual integrity of reconstructed images. Experimental results demonstrate that the proposed UDOR, despite unsupervised, achieves truly encouraging results on par with those of supervised methods.

To solve the problem of insufficient utilization of non-motorized lanes and traffic congestion of motor lanes in the cold areas of northern China, priority lanes for non-motorized vehicles are studied. First, the traffic status of the studied sections and intersections is evaluated through traffic investigations. Then, the non-motorized lanes of the basic sections and intersections are transformed into non-motorized priority lanes through the transformation of signs and markings. Finally, the effect of the transformation is evaluated by calculation and simulation methods. The results of the calculations and simulation show that the capacity of the road sections and intersections is improved by varying degrees, the saturation is been reduced, and the traffic situation is improved significantly. Therefore, in the cold areas of the north, for sections and intersections with traffic congestion, non-motorized lanes can be transformed into non-motorized priority lanes, which can not only alleviate traffic congestion but also improve the utilization of non-motorized lanes.

Satellite hardware has reached a level of development that enables imaging satellites to realize applications in the area of meteorology and environmental monitoring. As the requirements in terms of feasibility and the actual profit achieved by satellite applications increase, we need to comprehensively consider the actual status, constraints, unpredictable information, and complicated requirements. The management of this complex information and the allocation of satellite resources to realize image acquisition have become essential for enhancing the efficiency of satellite instrumentation. In view of this, we designed a satellite auto mission planning system, which includes two sub-systems: the imaging satellite itself and the ground base, and these systems would then collaborate to process complicated missions: the satellite mainly focuses on mission planning and functions according to actual parameters, whereas the ground base provides auxiliary information, management, and control. Based on the requirements analysis, we have devised the application scenarios, main module, and key techniques. Comparison of the simulation results of the system, confirmed the feasibility and optimization efficiency of the system framework, which also stimulates new thinking for the method of monitoring environment and design of mission planning systems.

At present, under the pressure of environmental pollution, logistics enterprises are beginning to use electric vehicles for various distribution services due to their low energy consumption and environmentally friendly nature. Considering the fact that the quality of service and charging strategy have an important impact on the electric vehicle routing problem, to improve the efficiency of electric vehicles in logistics distribution networks, we investigate an electric vehicle routing problem with soft time windows and fast charging (EVRPSTW-FC) stations: mixed integer linear programming is established to minimise total logistics energy consumption. To solve the proposed model, a hybrid adaptive genetic algorithm (HAGA) is proposed. The performance of HAGA is compared and tested with benchmark examples, and the results verify the feasibility and effectiveness of the proposed model and solution algorithm.

Satellite range scheduling plays a very important role in guaranteeing the normal operation and the real-time control of in-orbit satellites. Although there appears a stronger demand for multi-objective optimization of satellite monitoring departments, multiple scheduling criteria are rarely considered simultaneously. To address the multi-objective satellite range scheduling problem (MOSRSP), a general MOEA based memetic algorithm (MOEA-MA) framework is proposed, which optimizes the failure rate of ground-satellite communication requests and the load-balance degree of remote-tracking antennas. Based on a novel decision model for MOSRSP, the conflict-resolution and load-balance operators and the tabu search metaheuristic are designed to implement the local search operations in the MOEA-MA. Different types of the MOEAs, including the domination-based MOEAs, decomposition-based MOEAs and metric-based MOEAs are adopted to implement the evolutionary operations in the MOEA-MA. The highlight of this paper is the effective application of the MOEA-MAs to practical scheduling problems, where the two most concerning objectives are well addressed. The MOEA-MAs that adopt five well-known MOEAs are given and examined by the Benchmarks problems. Computational results indicate that the MOEA-MAs outperform the original MOEAs in terms of the metrics of coverage, hypervolume and spacing, which show good performance and application prospect for the MOSRSP.

With the rapid development of the economy and the improvement of the urbanization level, our city is increasing traffic volume, in which the traffic problem is appearing gradually, especially in the important city or in the big city. This article mainly aims at the example of transformation of Harbin city Dongli roundabout based on the collection of data, carries on the analysis to the traffic situation and put forward ideas and strategies to improve the optimization of the intersection, optimization ideas and improving measures will make the traffic simulation. Finally, I hope the analysis of Harbin Dongli roundabout improvement examples can help other similar circular intersection improvement.

The wireless ad hoc distribution power communication networks (WAH-DPCNs) are particularly vulnerable due to its specific features. In this paper, we describe the new intrusion detection system called zone-based intrusion detection system (ZBIDS). This system is a distributed IDS, in which two levels of hierarchical structure is defined. The ZBIDS system can detect not only the simple intrusions without sequence characters by pre-classifier based anomaly detection but also the complicated intrusions with the special sequence characters by a Markov chain based anomaly detection. In addition, ZBIDS system realizes co-detection via the local and global aggregation-correlation engines. According to the simulation results, it can be seen that Markov chain based anomaly detection engine of ZBIDS has the very good intrusion detection capability

Energy efficiency is an important issue for mobile devices in ad hoc networks since they often have to rely on their batteries. In consideration that current mesh-based multicast routing protocols don't tend to form energy-efficient multicast infrastructure, we propose a new energy-efficient multicast routing protocol (E/sup 2/MRP) for mobile ad hoc networks. The two main characteristics of E/sup 2/MRP are 1) using in turn the criteria for minimum energy consumed per packet and minimum maximum node cost during the course of relaying group (RG) creation and maintenance, and 2) forming a graph-based multicast infrastructure instead of a tree-based one. For comparison, E/sup 2/MRP is implemented together with multicast incremental power (MIP) protocol and on-demand multicast routing protocol (ODMRP). When over half of nodes join multicast group and nodes move at higher speed, as we observe, MIP results in higher energy consumption for each packet than E/sup 2/MRP and ODMRP due to the fragile multicast tree. Compared to MIP and ODMRP, E/sup 2/MRP tremendously reduces the energy consumption rate of nodes and hence prolongs the lifetime of nodes and networks, especially when the size of multicast group is small and node mobility is low. Simulation results demonstrate significant performance improvement.