一种遗传搜索块匹配运动估计算法

运动估计是帧间视频编码中的关键技术,但现有的快速搜索算法中大都是次优算法,且易陷于局部极小点,针对此问题,提出了将一种遗传算法应用于块运动估计中的遗传搜索匹配估计算法（GSAME）,该方法把块运动向量作为遗传染色体,经过杂交、变异等操作,以便得到全局意义上的最优解,并与经典的全局搜索法和三步搜索法进行了比较,实验结果显示,该算法不仅有效地解决了局部极小问题,而且计算量也较少。     

基于改进型C-1BT变换的快速全局运动估计

针对存在单一运动目标视频序列的全局运动估计问题,提出一种结合C-1BT变换和自适应十字搜索法(ARPS)的快速块匹配运动估计方法。采用简单的预处理,提高灰度突然变化和存在局部运动目标时的全局运动估计的鲁棒性;简化传统C-1BT变换算法中的核函数,降低计算负担;考虑到实际视频中存在大量的零运动矢量(ZMV)区域和视频的空间相关性,改进原有的ARPS搜索法。MATLAB仿真结果证明,提出的算法在保证精度的同时极大地减少了搜索点数。     

一种基于新型遗传算法的块运动估计算法

提出了一种基于新型遗传算法的块匹配运动估计算法。该算法把块运动向量作为遗传染色体,经过选择变异等操作,将随机搜索与特定目标搜索相结合,解决了以往快速搜索算法易陷于局部最优的问题,同时该算法中所带的模糊评价函数使得对个体的评价更合理、客观,该算法还将运动矢量空间偏置特性用于初始种群的选取,进一步提高了算法性能。实验结果表明,该算法性能上接近于FSA,速度却接近于TSS。     

基于时空相关性的自适应运动估计快速算法

针对视频编码中的核心技术运动估计,提出一种基于运动矢量特性的运动估计快速算法。算法分析视频序列运动矢量的特性,对静止块设定自适应阈值直接终止搜索,自适应选择搜索起点。采用2种模式进行搜索,依据当前块与邻近块的空间相关性大小,即局部运动性是高还是低,自适应地选择搜索模式和搜索终止门限。实验结果表明,该算法的搜索速度和搜索精度优于现有的快速运动估计算法,搜索精度非常接近于全搜索法。     

Fast global motion estimation based on improved C-1BT transformation

针对存在单一运动目标视频序列的全局运动估计问题, 提出一种结合C-1BT变换和自适应十字搜索法ARPS的快速块匹配运动估计方法。采用简单的预处理, 提高灰度突然变化和存在局部运动目标时的全局运动估计的鲁棒性; 简化传统C-1BT变换算法中的核函数, 降低计算负担; 考虑到实际视频中存在大量的零运动矢量ZMV区域和视频的空间相关性, 改进原有的ARPS搜索法。MATLAB仿真结果证明, 提出的算法在保证精度的同时极大地减少了搜索点数。     

一种综合搜索策略的快速运动估计算法*

提出了一种综合搜索策略的运动估计算法。该算法首先采用中值预测提前终止判断策略,然后基于块运动类型确定搜索起点,最后采用小十字模板与基于块的梯度下降搜索法(BBGDS)相结合的方法进行局部搜索。搜索过程中多处引入提前终止策略,进一步提高搜索速度。通过与综合性能代表当前国际先进水平的运动矢量场自适应搜索法(MVFAST)进行对比实验发现,该算法在基本保持搜索精度的情况下,有效提高了搜索速度,对于运动较大序列速度提高尤为明显,可以达到20%48%。     

扩散遗传算法及其应用研究

针对遗传算法在局部搜索能力方面的缺陷,提出了一种基于扩散算子的遗产算法(简称扩散遗产算法)。该算法中包含的扩散算子是变异算子,其主要作用是在遗传搜索中进行局部搜索。用扩散遗传算法和实数编码遗传算法分别训练用于解XOR问题的神经网络,对比结果表明,论文提出的算法兼具强的全局搜索能力和局部搜索能力,因此,该算法可以不借助其它局部搜索算法而单独作为神经网络训练算法,从而简化训练算法,提高训练效率。该算法对提高遗传算法搜索效率和求解精度具有重要的意义。     

一种快速分类搜索运动估计新方法

在视频编码中 ,基于块的运动估计算法被广泛应用 .在保证估计质量的前提下 ,为了降低运动估计算法的搜索次数 ,提出了一种对于不同类型的块采用不同的搜索范围和搜索步骤的分类快速搜索 (CFS)运动估计新算法 .该算法首先对块进行分类 ,然后确定其搜索范围和搜索步骤 ,在应用分类搜索法时 ,根据运动矢量的中心偏置特性 ,将第 1步和第 2步的搜索窗采用 5× 5的窗口 ,第 3步采用 3× 3的窗口 .结果表明 ,该分类快速搜索新算法在运动矢量的估计质量上 ,明显优于传统三步搜索法 ,且搜索次数与传统三步搜索法相比 ,降低了 2 3% ,与全搜索法相比 ,降低了 91% .实验结果证明 ,该算法尤其适用于快速运动、复杂运动序列的运动估计 .与传统的全搜索法和三步搜索法相比 ,其更适合于用硬件实现 .     

一种改进的快速全局运动估计算法

结合两步法与传统梯度下降算法,提出一种改进的快速全局运动估计算法。采用稀疏抽样的MSEA快速块匹配算法估计局部运动矢量,使用迭代最小二乘法粗估计全局运动参数并排除外点(前景宏块),在排除外点的采样宏块集上选取特征像素,以上述两步法的全局运动估计参数为初始值,利用LM梯度下降算法对全局运动参数进行优化。实验结果表明,改进算法的估计速度达到11.42 ms/f,比FFRGMET算法快1.3倍,具有更高的全局运动估计精度。     

基于遗传模拟退火算法的门阵列布局方法

为实现门阵列模式布局,将遗传算法与模拟退火算法相结合,提出一种新的遗传模拟退火算法,利用遗传算法进行全局搜索,利用模拟退火法进行局部搜索,在进化过程中采用精英保留策略,对进化结果进行有选择的模拟退火操作,既加强了局部搜索能力又防止陷入局部最优。实验结果表明,与传统遗传算法相比,该算法能够有效提高全局搜索能力。     

Simple and efficient motion estimation algorithm by continuum search

Motion estimation plays a vital role in reducing temporal correlation in video codecs but it requires high computational complexity. Different algorithms have tried to reduce this complexity. However these reduced-complexity routines are not as regular as the full search algorithm (FSA). Also, regularity of an algorithm is very important in order to have a hardware implementation of that algorithm even if it leads to more complexity burden. The goal of this paper is to develop an efficient and regular algorithm which mimics FSA by searching a small area exhaustively. Our proposed algorithm is designed based on two observations. The first observation is that the motion vector of a block falls within a specific rectangular area designated by the prediction vectors. The second observation is that in most cases, this rectangular area is smaller than one fourth of the FSA’s search area. Therefore, the search area of the proposed method is adaptively found for each block of a frame. To find the search area, the temporal and spatial correlations among motion vectors of blocks are exploited. Based on these correlations, a rectangular search area is determined and the best matching block in this area is selected. The proposed algorithm is similar to FSA in terms of regularity but requires less computational complexity due to its smaller search area. Also, the suggested algorithm is as simple as FSA in terms of implementation and is comparable with many of the existing fast search algorithms. Simulation results show the claimed performance and efficiency of the algorithm.     

Max-min surrogate-assisted evolutionary algorithm for robust design

Solving design optimization problems using evolutionary algorithms has always been perceived as finding the optimal solution over the entire search space. However, the global optima may not always be the most desirable solution in many real-world engineering design problems. In practice, if the global optimal solution is very sensitive to uncertainties, for example, small changes in design variables or operating conditions, then it may not be appropriate to use this highly sensitive solution. In this paper, we focus on combining evolutionary algorithms with function approximation techniques for robust design. In particular, we investigate the application of robust genetic algorithms to problems with high dimensions. Subsequently, we present a novel evolutionary algorithm based on the combination of a max-min optimization strategy with a Baldwinian trust-region framework employing local surrogate models for reducing the computational cost associated with robust design problems. Empirical results are presented for synthetic test functions and aerodynamic shape design problems to demonstrate that the proposed algorithm converges to robust optimum designs on a limited computational budget.     

基于改进遗传算法的稀疏重构算法

稀疏重构算法中凸松弛法在恢复效率方面、贪婪追踪法在恢复精度方面存在不足,基于遗传算法迭代优化的思想,结合模拟退火以及多种群算法的优势,提出了基于模拟退火遗传算法和基于多种群遗传算法的启发式稀疏重构算法。所提算法均从传统遗传算法易陷入局部最优解的缺陷出发,分别通过保持个体间的差异性和提高种群多样性来搜索待求稀疏信号的全局最优解,并通过理论分析证明了所提算法参数选取及搜索策略的有效性。此外,以阵列信号处理中空间信源的波达方向（DOA）估计问题为例,验证所提算法的有效性。仿真结果表明,相较于正交匹配追踪OMP算法和基于l1范数奇异值分解的l1-SVD算法,所提算法提高了DOA估计的精度,且降低了运算复杂度,使其快速收敛至全局最优解。     

Fast simulated annealing hybridized with quenching for solving job shop scheduling problem

Various heuristic based methods are available in literature for optimally solving job shop scheduling problems (JSSP). In this research work a novel approach is proposed which hybridizes fast simulated annealing (FSA) with quenching. The proposed algorithm uses FSA for global search and quenching for localized search in neighborhood of current solution, while tabu list is used to restrict search from revisiting previously explored solutions. FSA is started with a relatively higher temperature and as search progresses temperature is gradually reduced to a value close to zero. The overall best solution (BS) is maintained throughout execution of the algorithm. If no improvement is observed in BS for certain number of iterations then quenching cycle is invoked. During quenching cycle current temperature is reduced to nearly freezing point and iterations are increased by many folds, as a result of this change search becomes nearly greedy. The strength of the proposed algorithm is that even in quenching mode escape from local optima is possible due to use of Cauchy probability distribution and non-zero temperature. At the completion of quenching cycle previous values of search parameters are restored and FSA takes over, which moves search into another region of solution space. Effectiveness of proposed algorithm is established by solving 88 well known benchmark problems taken form published work. The proposed algorithm was able to solve 45 problems optimally to their respective best known values in reasonable time. The proposed algorithm has been compared with 18 other published works. The experimental results show that the proposed algorithm is efficient in finding solution to JSSP.     

全局调距和声特征选择算法

特征选择技术能有效解决维数灾难问题，许多搜索策略已经被应用到特征选择问题中。针对和声特征选择算法搜索能力低下的问题，提出了一种基于全局自适应调距的和声特征选择算法（HSFS-GPA）。将特征集的距离定义引入到特征选择问题中，在算法搜索过程中结合全局信息对随机产生的新和声进行调整，以一定概率减小候选和声与当前最优和声的距离来加快算法搜索速度，或减少候选和声与最差和声的距离以避免陷入局部最优；同时，采用竞争选择方案随时更新和声库全局信息，改进和声库的更新机制提高算法搜索质量。将HSFS-GPA与原始和声特征选择算法、粒子群算法和遗传算法进行对比实验，HSFS-GPA所选特征子集的大小比原始和声算法减少15%，子集评价值平均提高到0.98。实验结果表明，HSFS-GPA能在相同的条件下搜索到更优质的特征子集。     

An improved artificial bee colony algorithm based on the strategy of global reconnaissance

The artificial bee colony (ABC) algorithm is a recently introduced swarm intelligence optimization algorithm based on the foraging behavior of a honeybee colony. However, many problems are encountered in the ABC algorithm, such as premature convergence and low solution precision. Moreover, it can easily become stuck at local optima. The scout bees start to search for food sources randomly and then they share nectar information with other bees. Thus, this paper proposes a global reconnaissance foraging swarm optimization algorithm that mimics the intelligent foraging behavior of scouts in nature. First, under the new scouting search strategies, the scouts conduct global reconnaissance around the assigned subspace, which is effective to avoid premature convergence and local optima. Second, the scouts guide other bees to search in the neighborhood by applying heuristic information about global reconnaissance. The cooperation between the honeybees will contribute to the improvement of optimization performance and solution precision. Finally, the prediction and selection mechanism is adopted to further modify the search strategies of the employed bees and onlookers. Therefore, the search performance in the neighborhood of the local optimal solution is enhanced. The experimental results conducted on 52 typical test functions show that the proposed algorithm is more effective in avoiding premature convergence and improving solution precision compared with some other ABCs and several state-of-the-art algorithms. Moreover, this algorithm is suitable for optimizing high-dimensional space optimization problems, with very satisfactory outcomes.     

一种新的非线性回归模型参数估计算法

提出一种新的基于混合基因算法（HGA）的非线性回归模型参数估计算法,新算法通过对问题的解空间交替进行全局和局部搜索,达到快速收敛至全局最优解,较好地解决了传统算法通用性差、易陷入局部极小的问题,实验验证了算法的通用性和有效性。     

一种基于云模型的改进型量子遗传算法*

针对量子遗传算法在函数优化中易陷入局部最优和早熟收敛等缺点,采用云模型对其进行改进,采用量子种群基因云对种群进化进行定性控制,采用基于云模型的量子旋转门自适应调整策略进行更新操作,使算法在定性知识的指导下能够自适应控制搜索空间范围,能在较大搜索空间条件下避开局部最优解。典型函数对比实验表明,该算法可以避免陷入局部最优解,能提高全局寻优能力,同时能以更快的速度收敛于全局最优解,优化质量和效率都要优于遗传算法和量子遗传算法。