Abstract:

Abstract: Abstract: A model independent attitude control method based on genetic algorithm(GA)was proposed to solve the high-precision attitude control problem of complex and nonlinear uncertain systems such as variable mass charactenstic spacecraft. The algorithm optimized the control parameters based on the indepen dent scale of the spacecraft model, without relying on the quality parameter identification of the spacecraft. The simulation results showed that the model independent attitude control method based on genetic algorithm could control the spacecraft with variable mass characteristics under the condition of controlling torque and angular velocity constraints. The algorithm would be of great value for the use of orbiting spacecraft.

Abstract: For solving the dynamic economic emission dispatch problem (DEED), a multiple learning based multi-objective pigeon-inspired optimization (MLMPIO) algorithm is proposed in this paper. In the proposed multiple learning strategy, individuals of the population are allowed to learn from multiple global best positions of the external archive and from the personal historical best positions. This learning strategy enables the preservation of diversity and global search ability of the population to prevent premature convergence. Meanwhile, small probability mutation is introduced to MLMPIO to enhance the swarm diversity and search ability further. The external archive with adaptive changing capacity is used to store the current Pareto optimal solutions. To verify the performance of the proposed method, the DEED problem of the IEEE 10-generator power system has been solved. And the results demonstrate the feasibility and effectiveness of the proposed method

Abstract: undefined

In this paper, a trajectory planning approach based on the principle of dynamie programming andframe work of pigeon inspired optimization( PIO) was proposed for UAV surveillance tasks. In this approachthe sensor visibility was firstly analyzed by considering the occ lusions caused by terrain feature, and the delee.table areas of the targets were approximated by a series of poly gons. To determine the optimal trackable path togons were replaced by with their centers firstly, which allowed tocover all target sites, the target visibility poly obtained an initial solution by optimizing the order of the targets to be visited. In the following step of the algorithm, a path refinement scheme combing dynamic programming and PIO was proposed to refine the initialconsidering t he sensor visibility and turning radius constraint of the UAV. Comparative simulationroute

Abstract: In this paper,an evolutionary game theory based pigeon-inspired optimization (EGTPIO) algorithm is proposed to improve the performance of the deformable ground mobile robot (DGMR) by using active disturbance rejection control (ADRC) to control the executing mechanism in DGMR. EGPIO not only keeps the advantage of fast convergence of PIO, but also determines the proportion of the two operators through the process of group evolution. This reduces the probability of results falling into local optimum and improves the speed and stability of ADRC controller. The simulation results show that the superiority of EGPIO in the quantity, quality and convergence speed of optimal solutions makes ADRC reach stable state quickly and improves the maneuverability of DGMR

Abstract: Pigeon inspired optimization (PIO) algorithm , as an emerging optimization technology, has the advantages of fast convergence and high precision. But it is not ideal for some problems with local optimal values. By introducing lost&exploration and c luster splitting mechanisms of natural flying pigeons, an improved PIO algorithm based on lost&exploration and cluster splitting (LSPIO) is proposed in this paper. The lost&exploration mechanism enhances the global search performance of the algorithm, and the cluster splitting mechanism increases the diversity of the population. In this paper, 9 standard test functions are selected for algorithm performance evaluation. Compared with standard pigeon group algorithm and particle swarm algorithm, the results show that the new LSPIO algorithm can effectively avoid premature problems while maintaining good convergence properties

Abstract: In this paper, a new multi-focus image fusion algorithm is proposed based on convolution neural network in non-subsampled Shearlet (NSST) domain by using the advantages of time-frequency of NSST. Firstly, the source image is decomposed by NSST. Secondly, the fusion strategy based on the convolution neural network (CNN) is applied to the low frequency coefficients of the decomposition. Then, the improved weighted sum of Laplace energy based on the guided filtering are carried out to the high-frequency coefficients of the decomposition. Finally, the fused image can be gotten by inverse NSST transform. The algorithm fully preserves the information of the source image and improves the continuity of the image space. Experimental results show that the fusion algorithm can not only achieve better visual effects, but also improve its objective evaluation index.

Abstract: Image segmentation is a kind of constrained nonlinear optimization problem that needs to seek the optimal solution in nonlinear parameter space. In order to improve the precision of the optimization problem, an image segmentation method based on pigeon group optimization algorithm is proposed. First, the segmentation threshold is used as the optimization variable, and the image segmentation is modeled as a nonlinear optimization problem with the optimal threshold equation as the objective function, and the inter-class variance and the w 0 and w1 ranges as the constraints. Then, Using random segmentation threshold as the initial value of iteration, the optimal parameters are solved by the pigeon group optimization algorithm(PIO). In order to verify the validity of this method, two kinds of images with different features are divided into experiments, and the algorithm is evaluated by overlapping degree and time efficiency. The results show that the algorithm has the highest degree of overlap and the shortest operation time.

Abstract: In order to solve the problem of shared bicycle scheduling, a discrete differential evolution algorithm is designed to solve it. This paper systematically introduces the solution of discrete differential evolution algorithm in the face of shared bicycle scheduling problem, improves its mutation operator and repair operator for scheduling problem, and proposes a new mutation and repair method for discrete variables and path planning problem. The specific path can be calculated in the execution process. The results show that compared with greedy algorithm and ant colony algorithm, the proposed algorithm has higher quality and faster convergence speed and has certain practical value in a series of scheduling problems such as shared bicycle scheduling.

Abstract: Under the assumption that the total amount of plants is constant, the locations of plants on the island become more and more concentrated at the highest point with the rising of sea level. Inspired by this phenomenon, a metaheuristic algorithm, Island algorithm (IA), is proposed. The Island algorithm consists of three phases in each iteration: elimination phase, sea level rising phase and balance phase. By analyzing the IA algorithm, the reason of the algorithm advantage, the characteristics of the algorithm and the characteristics of the suitable and unsuitable functions are found out. The complexity and robustness of IA algorithm are analyzed.. The algorithm is applied to eight typical test functions in many dimensions and compared with the PSO algorithm and BA algorithm. The results show that IA algorithm is slightly worse than PSO algorithm and BA algorithm in calculating functions with certain characteristics; In the results of the other test functions, the accuracy and robustness of the algorithm are significantly better than PSO algorithm and BA algorithm in many dimensions, which verifies the analysis of IA algorithm.

Abstract: In this paper, an image encryption scheme based on chaos system and genetic operation is proposed. First, the sha-3 algorithm is used to calculate the hash value of the plain-text image as the initial value of the chaotic system. Secondly, using the sensitivity of chaotic map to initial conditions and pseudo-randomness, the pseudo-random sequence generated by Logistic map generates hill matrix, and the image is scrambled and replaced. Thirdly, combining Duffing mapping and DNA coding technology, pixel selection, crossover and mutation can be realized at the level of genetic manipulation to complete pixel diffusion and scrambling, which significantly increases the difficulty of decoding the algorithm. Finally, the confusion and diffusion characteristics of the algorithm are further enhanced by bidirectional XOR with chaotic sequences. Experimental and security analysis results show that the algorithm is sensitive to key and can effectively resist statistical attack and differential attack. It has good security and application potential. The image encryption effect and performance are improved significantly.

Abstract: In the human-computer interaction system based on fingertip, the position of fingertip center is very important. By solving the multi-objective optimization model for the fingertip localization, several fingertip center positions can be obtained. The fingertip pixels distribute around the fingertip centers, so the optimal solution components of this optimization model have the above distribution law. Using the estimation of distribution algorithm with the distribution law to solve this optimization model, can obtain accurate results. This paper discusses the estimation of distribution algorithm for the fingertip localization. It proposes that the decision variable dimension, population size, maximum sampling variance, and minimum sampling variance are the main parameters of this estimation of distribution algorithm. The experimental results show that each main parameter has its best value; when their values are their best values, the fingertip center positions obtained by the proposed method excel the results of the existing methods.

Abstract: In order to solve the state explosion problem caused by the coupling of fuzzy causality and inaccurate information in complex system structure, this paper presents a hierarchical decoupling method based on ODDT for FDES composite causal chain.. The method first constructs a Petri nets model (TC-PPN) for decoupling the causal chain under the time constraints. Then based on the merged state information and timing information, the conception and measurement method of Observable Degree in Dimensionality of Time (ODDT) of complex systems in FDES are further proposed by constructing the time constrained graph of observation information.Finally, based on the time information of the global state, the degree of observability is calculated and a hierarchical decoupling method based on ODDT is proposed

Abstract: Urban capital structure was a complex?problem affected by multi-factors and multi-objective particle.This paper attempt ed to explore a scientific and appropriate d algorithm to construct the optimal capital structure model under the influence of multi-objective and multi-factors to analyze the situation of urban capital structure.First, the data in history could find the relationship among features of the data in history by using the regression characteristics of random forest. Then, the multi-objective particle swarm optimization algorithm was used to find values of the features that achieve the best results according to the existing relationship features. Then finding the most correlate data from the historical data based on the best eigenvalues of these effects. Therefore, the cities and the years with relatively better capital structure allocations are analyzed. We could play a good role in the reference and development of each city by continuously learning these superior structural configurations

Abstract: First, according to the theory of space crowdsourcing, the concept of equivalent task representative points is proposed, and the relationship between the original task pricing law and task density, membership density, member average credibility and nearest neighbor reach distance is studied. On this basis, from the perspectives of the contractor, the platform and the contractor, According to the four steps of completing the task, a task pricing model based on multi-objective programming , a member dynamic grab order model, a task allocation model and a task completion probability prediction model are respectively established. Furthermore, the TOPSIS method is used to calculate the comprehensive evaluation index of different pricing schemes, and then choose the optimal task pricing scheme by the ranking result of the comprehensive evaluation index. Finally, the optimized scheme is compared with the original scheme. Under the condition that the total cost of the contractor is as low as possible, the platform task completion rate, the average individual member income and the unit reputation value conversion rewards are significantly improved, that is, the crowdsourcing performance are improved. The result verifies the feasibility and effectiveness of the model and provides reference for the task pricing of the crowdsourcing platform

Abstract: A new type of coupling spiral baffles heat exchangers is presented coupling the ideal spiral baffle with external thread tube in the shellside. With the CFD code FLUENT, Analysing the shell side heat transfer mechanism under coupling screw action, and compared with that of ideal helical baffle heat exchanger. The mechanism of heat transfer enhancement in shellside was analyzed. It is shown that the comprehensive performance of shellside in coupling spiral baffle heat exchanger increases by 4.5%~14.5% than that of ideal helical baffle heat exchanger, when the helical angle are 10°, 15° and 20° ，and shell side Reynolds number ranges 2000 to 6000. When the helical angle is 10° and Reynolds number is 4000, the heat transfer coefficient, comprehensive performance, temperature field and pressure field in shellside of coupling spiral heat exchanger are all better than that of the ideal spiral baffle heat exchanger. The study and results provide references for the improvement of heat transfer of helical baffles heat exchanger

Abstract: Aiming at the problem of predicting the future demand of water resources in Nanchang, a water resource demand forecasting method based on particle swarm optimization algorithm is proposed. Based on the historical population, economy and water demand data of Nanchang City, linear, exponential and mixed forecasting models are constructed. The algorithm optimizes the prediction model to determine the model parameters. The simulation experiment results show that all three models can obtain good prediction accuracy, and the hybrid prediction model is the best, with a prediction accuracy of 97.71%.

Abstract: