共查询到19条相似文献,搜索用时 156 毫秒
1.
动态时间规整算法DTW(Dynamic Time Warping)作为一种非线性时间匹配技术已成功地应用于语音识别系统中。DTW算法使用动态规划技术来搜索两个时间序列的最优规整路径,虽然这种算法计算量小,运算时间较短,但只是一种局部优化算法。禁止搜索TS(Tabu Search)算法是一种具有短期记忆的广义启发式全局搜索技术,适用于解决许多非线性优化问题。本文将该技术用于语音识别系统中,提出了基于禁止搜索的非线性时间规整的优化算法TSTW,使得时间规整函数尽可能逼近全局最优。仿真结果表明,TSTW比DTW有更高的识别率,且运行时间比遗传时间规整算法GTW大大减少。 相似文献
2.
基于轮廓的图像匹配是计算机视觉领域中的重要问题,但是目前尚未有较成熟的算法能够很好地解决局部轮廓匹配问题及非相似变换和非刚体变换引起的轮廓形变问题.根据局部轮廓结构在产生形变时具有相对稳定性的规律及融合轮廓局部信息和全局信息的轮廓描述思想,本文提出了一种具有尺度、旋转、平移不变性,形变鲁棒性和初始点无关性的局部尺度轮廓描述算法.在此基础上,针对线性匹配方法效果不佳以及传统DTW技术约束路径的线性度不满足轮廓采样特性要求的问题,提出一种基于改进DTW技术的轮廓匹配算法,即结合轮廓采样特性设置九宫格的路径约束条件,以旋转角度为参数,计算全局最佳匹配路径.实验结果表明,对于存在尺度、平移、旋转及形变关系的两轮廓,该方法能较好地实现轮廓间的局部匹配,并且其匹配准确率平均约为92%,较HD算法提高了30%,较传统DTW算法提高了26%. 相似文献
3.
4.
5.
6.
7.
8.
9.
10.
基于改进蚁群算法的机器人路径规划 总被引:1,自引:0,他引:1
采用MAKLINK图论建立机器人路径规划的空间模型,利用Dijkstra算法减少工作空间的搜索范围,引入免疫算子,将其融合到蚁群算法的每次迭代过程中,提高蚁群算法在全局搜索空间的遍历性和收敛速率,避免陷入局部最优解。 相似文献
11.
12.
为了解决网络层析成像中链路故障诊断的NP难问题,提出一种基于蚁群算法的故障链路诊断方法。首先将问题建模成一个组合优化问题,利用蚁群算法在解决组合优化问题中独特的优势进行求解。不同于传统的蚁群算法,求解故障链路时蚁群在初始放置点和可行路径上都受约束。为了加快算法的收敛速度,对蚁群算法的初始信息素浓度进行优化。仿真结果表明,所提出的算法在故障链路检测中具有较好的精度和召回率。 相似文献
13.
In this paper, a Tabu search based routing algorithm is proposed to efficiently determine an optimal path from a source to a destination in wireless sensor networks (WSNs). There have been several methods proposed for routing algorithms in wireless sensor networks. In this paper, the Tabu search method is exploited for routing in WSNs from a new point of view. In this algorithm (TSRA), a new move and neighborhood search method is designed to integrate energy consumption and hop counts into routing choice. The proposed algorithm is compared with some of the ant colony optimization based routing algorithms, such as traditional ant colony algorithm, ant colony optimization-based location-aware routing for wireless sensor networks, and energy and path aware ant colony algorithm for routing of wireless sensor networks, in term of routing cost, energy consumption and network lifetime. Simulation results, for various random generated networks, demonstrate that the TSRA, obtains more balanced transmission among the node, reduces the energy consumption and cost of the routing, and extends the network lifetime. 相似文献
14.
Aiming at the multi-constraint routing problem,a mathematical model was designed,and an improved immune clonal shuffled frog leaping algorithm (IICSFLA) was proposed,which combined immune operator with traditional SFLA.Under the constraints of bandwidth,delay,packet loss rate,delay jitter and energy cost,total energy cost from the source node to the terminal node was computed.The proposed algorithm was used to find an optimal route with minimum energy cost.In the simulation,the performance of IICSFLA with adaptive genetic algorithm and adaptive ant colony optimization algorithm was compared.Experimental results show that IICSFLA solves the problem of multi-constraints QoS unicast routing optimization.The proposed algorithm avoids local optimum and effectively reduces energy loss of data on the transmission path in comparison with adaptive genetic algorithm and adaptive ant colony optimization algorithm. 相似文献
15.
针对蚁群算法易陷入局部最优、收敛速度慢的问题,文中提出了一种基于拥挤度因子的动态信息素更新策略的蚁群算法(CFACS)。引入鱼群算法中拥挤度的思想,扩大种群中蚂蚁分布范围,使其探索更大的解空间,提高算法全局搜索能力;采用动态信息素更新策略,在每一次迭代中,自适应调整当前最优路径所释放的信息素浓度,保证蚁群前期的多样性,同时保证算法在后期的收敛性。求解TSP问题的仿真实验表明,改进算法求得解的质量和求解的收敛速度都明显优于传统蚁群算法,较好地平衡了种群多样性与收敛速度之间的矛盾。 相似文献
16.
莫桂江 《微电子学与计算机》2011,28(9)
为了提高无线传感器网络路径优化效率,快速找到最优路径,提出基于蚁群-遗传算法的传感器路径优化方法.利用遗传算法的快速全局搜索能力和蚁群算法的正反馈机制,实现了两种算法的融合.仿真结果表明,蚁群-遗传算法在时间和性能上都优于单独的蚁群算法和遗传算法,能快速找到无线传感器网络最优路径,有效延长了网络的生命周期. 相似文献
17.
针对传统的光纤光栅电压传感器非线性校正算法具 有运行速度慢,拟合精度不高的缺陷。在研究了大量国内外文献过后,本文为了解决一些传 统非线性校正方法在光栅光纤传感器校正中的不足,在此提出了一种基于蚁群算法优化的分 段支持向量机回归的 校正算法。由于传统的蚁群算法在信号处理中搜索速度不理想,最小二乘支持向量机回归算 法精度不高,所以此算法是结合了蚁群 算法搜索最小二乘支持向量机回归最佳参数原理的基础上将样本空间按照数据分布情况进行 分段回归,以此减少算法运行时间。首 先通过蚁群算法优化各个支持向量机参数,然后通过分段回归得到传感器完整的特性,曲线 拟合精度为99.97%。此算法克服了传统 支持向量机回归算法中局部最优解的问题,具有较好的全局收敛效果。 相似文献
18.
Forthe problem that in interactive network,the illegal and abnormal behaviors were becoming more hidden,moreover,the complex relation in real interactive network heightens the difficulty of detecting anomalous entities,an ant colony model was proposed for extracting the backbone network from the complex interactive network.The novel model simulated the relationships among entities based on the theory of path optimization,reduced the network size after quantifying the significance of each flow of information.Firstly,a strategy of initial location selection was proposed taking advantage of network centrality.Secondly,a novel path transfer mechanism was devised for the ant colony to fit the flow behavior of entities.Finally,an adaptive and dynamic pheromone update mechanism was designed for guiding the optimization of information flows.The experimental results show that the proposed model is superior to the traditional ant colony algorithm in both solving quality and solving performance,and has better coverage and accuracy than the greedy algorithm. 相似文献