公开(公告)号：US07577548B1

公开(公告)日：2009-08-18

申请号：US11713561

申请日：2007-03-01

申请人： Krzysztof W Przytula ,  Shubha Kadambe ,  Narayan Srinivasa

发明人： Krzysztof W Przytula ,  Shubha Kadambe ,  Narayan Srinivasa

IPC分类号： G06F19/00 ,  G06F17/40

CPC分类号： G06N7/005

摘要： Described is a system for diagnosis and prognosis of a component. The system is configured to receive a signal from a component. The signal is representative of a current health observation of the component. The system also computes a present likelihood of the component failure based on the signal. Additionally, the system computes a future likelihood of failure of the component for a given future mission. Through diagnosis, a user can determine the present health of the component, and based on the present health and future mission, determine whether or not the component will fail in the future mission.

摘要翻译： 描述了用于组件的诊断和预后的系统。 系统被配置为从组件接收信号。 该信号代表组件的当前健康观察。 该系统还基于该信号计算组件故障的当前可能性。 此外，该系统计算组件在给定未来任务中的未来可能性。 通过诊断，用户可以确定组件的当前健康状况，并根据当前健康状况和未来任务，确定组件是否在将来的任务中失败。

公开(公告)号：US20080256009A1

公开(公告)日：2008-10-16

申请号：US11786949

申请日：2007-04-12

申请人： Qin Jiang ,  Narayan Srinivasa

发明人： Qin Jiang ,  Narayan Srinivasa

IPC分类号： G06N3/02

CPC分类号： G06N3/0436

摘要： Described is a system for temporal prediction. The system includes an extraction module, a mapping module, and a prediction module. The extraction module is configured to receive X(1), . . . X(n) historical samples of a time series and utilize a genetic algorithm to extract deterministic features in the time series. The mapping module is configured to receive the deterministic features and utilize a learning algorithm to map the deterministic features to a predicted {circumflex over (x)}(n+1) sample of the time series. Finally, the prediction module is configured to utilize a cascaded computing structure having k levels of prediction to generate a predicted {circumflex over (x)}(n+k) sample. The predicted {circumflex over (x)}(n+k) sample is a final temporal prediction for k future samples.

摘要翻译： 描述了一种用于时间预测的系统。 该系统包括提取模块，映射模块和预测模块。 提取模块被配置为接收X（1），。 。 。 X（n）时间序列的历史样本，并利用遗传算法提取时间序列中的确定性特征。 映射模块被配置为接收确定性特征并利用学习算法将确定性特征映射到时间序列的预测x（n + 1）样本。 最后，预测模块被配置为利用具有k级预测的级联计算结构来生成预测的x（n + k）样本。 预测的x（n + k）样本是k个未来样本的最终时间预测。

公开(公告)号：US20060129843A1

公开(公告)日：2006-06-15

申请号：US11198798

申请日：2005-08-05

申请人： Narayan Srinivasa ,  Swarup Medasani ,  Yuri Owechko ,  Deepak Khosla

发明人： Narayan Srinivasa ,  Swarup Medasani ,  Yuri Owechko ,  Deepak Khosla

IPC分类号： G06F12/14

CPC分类号： G06F16/335 ,  Y10S707/99933 ,  Y10S707/99935 ,  Y10S707/99942 ,  Y10S707/99943 ,  Y10S707/99945

摘要： An apparatus and method is disclosed for providing application specific multi-dimensional information to an application running on a user computing device, wherein at least one dimension of the information is a category, from a plurality of member documents electronically extracted from a library of electronically searchable documents, which may comprise an application specific multidimensional information extractor adapted to extract occurrences of prospective representations of dimensions of application specific multidimensional information from the member documents, and to extract occurrences of non-application specific multidimensional information from the member documents; and, an encoder adapted to encode the occurrences of prospective dimensions of application specific multidimensional information and non-application specific multidimensional information contained in member documents according to a dimension specific coded representation of each dimension of application specific multidimensional information and a non-application specific coded representation of each non-application specific multidimensional information element. The apparatus and method may further comprise a member document identifier adapted to determine whether a member document contains coded formatting, and if not, whether the member document is a dense document, and if not, for rejecting the document from further processing, and the coded formatting may comprise network markup language coding. The apparatus and method may further comprise an application specific multidimensional information verification unit adapted verify the extraction of application specific multi-dimensional information from the member documents, and may further comprise a database for storing the application specific multi-dimensional information adapted to provide an application running on a user computing device access to the application specific multidimensional information. The application specific multidimensional information may be scheduled events having the dimensions of time, location and event identity, and the application running on the user computer can be an electronic calendar or other similar scheduling software program.

公开(公告)号：US06456991B1

公开(公告)日：2002-09-24

申请号：US09388858

申请日：1999-09-01

申请人： Narayan Srinivasa ,  Yuri Owechko

发明人： Narayan Srinivasa ,  Yuri Owechko

IPC分类号： G06N302

CPC分类号： G06K9/6222 ,  G06N3/0409 ,  G06N3/0454 ,  G06N3/082

摘要： A boosting and pruning system and method for utilizing a plurality of neural networks, preferably those based on adaptive resonance theory (ART), in order to increase pattern classification accuracy is presented. The method utilizes a plurality of N randomly ordered copies of the input data, which is passed to a plurality of sets of booster networks. Each of the plurality of N randomly ordered copies of the input data is divided into a plurality of portions, preferably with an equal allocation of the data corresponding to each class for which recognition is desired. The plurality of portions is used to train the set of booster networks. The rules generated by the set of booster networks are then pruned in an intra-booster pruning step, which uses a pair-wise Fuzzy AND operation to determine rule overlap and to eliminate rules which are sufficiently similar. This process results in a set of intra-booster pruned booster networks. A similar pruning process is applied in an inter-booster pruning process, which eliminates rules from the intra-booster pruned networks with sufficient overlap. The final, derivative booster network captures the essence of the plurality of sets of booster networks and provides for higher classification accuracy than available using a single network.

摘要翻译： 提出了一种用于利用多个神经网络，优选基于自适应共振理论（ART）的神经网络的增强和修剪系统和方法，以便增加模式分类精度。 该方法利用输入数据的多个N个随机排列的副本，其被传递到多组增强网络。 输入数据的多个N个随机排列的副本中的每一个被分成多个部分，优选地具有与期望识别的每个类对应的数据的相等分配。 多个部分用于训练该组增强网络。 然后由增强器网络组生成的规则在促进器内修剪步骤中被修剪，其使用成对的模糊AND操作来确定规则重叠并消除足够相似的规则。 该过程导致一组内部加速器修剪的增强网络。 一个类似的修剪过程被应用在增强器之间的修剪过程中，该过程消除了具有足够重叠的内部加速器修剪的网络的规则。 最终的衍生增强网络捕获多组增强网络的本质，并提供比使用单个网络可用的更高的分类精度。

公开(公告)号：US20180308201A1

公开(公告)日：2018-10-25

申请号：US15494905

申请日：2017-04-24

申请人： Abhishek R. Appu ,  Altug Koker ,  Linda L. Hurd ,  Dukhwan Kim ,  Mike B. Macpherson ,  John C. Weast ,  Feng Chen ,  Farshad Akhbari ,  Narayan Srinivasa ,  Nadathur Rajagopalan Satish ,  Joydeep Ray ,  Ping T. Tang ,  Michael S. Strickland ,  Xiaoming Chen ,  Anbang Yao ,  Tatiana Shpeisman

发明人： Abhishek R. Appu ,  Altug Koker ,  Linda L. Hurd ,  Dukhwan Kim ,  Mike B. Macpherson ,  John C. Weast ,  Feng Chen ,  Farshad Akhbari ,  Narayan Srinivasa ,  Nadathur Rajagopalan Satish ,  Joydeep Ray ,  Ping T. Tang ,  Michael S. Strickland ,  Xiaoming Chen ,  Anbang Yao ,  Tatiana Shpeisman

IPC分类号： G06T1/20

CPC分类号： G06T1/20 ,  G06F9/3001 ,  G06F9/3017 ,  G06F9/3851 ,  G06F9/3887 ,  G06F9/3895 ,  G06F9/46 ,  G06N3/063 ,  G06T15/005 ,  G06T15/04 ,  G09G5/363

摘要： An apparatus to facilitate compute optimization is disclosed. The apparatus includes sorting logic to sort processing threads into thread groups based on bit depth of floating point thread operations.

公开(公告)号：US08977578B1

公开(公告)日：2015-03-10

申请号：US13535114

申请日：2012-06-27

申请人： Jose M. Cruz-Albrecht ,  Narayan Srinivasa ,  Peter Petre ,  Youngkwan Cho ,  Aleksey Nogin

发明人： Jose M. Cruz-Albrecht ,  Narayan Srinivasa ,  Peter Petre ,  Youngkwan Cho ,  Aleksey Nogin

IPC分类号： G06F15/18 ,  G06N3/04

CPC分类号： G06N3/04

摘要： A synaptic time-multiplexed (STM) neuromorphic network includes a neural fabric that includes nodes and switches to define inter-nodal connections between selected nodes of the neural fabric. The STM neuromorphic network further includes a neuromorphic controller to form subsets of a set of the inter-nodal connections representing a fully connected neural network. Each subset is formed during a different time slot of a plurality of time slots of a time multiplexing cycle of the STM neuromorphic network. In combination, the inter-nodal connection subsets implement the fully connected neural network. A method of synaptic time multiplexing a neuromorphic network includes providing the neural fabric and forming the subsets of the set of inter-nodal connections.

摘要翻译： 突触时间复用（STM）神经元网络包括神经织构，其包括用于定义神经织物的选定节点之间的节间连接的节点和开关。 STM神经元网络还包括神经形态控制器，以形成代表完全连接的神经网络的一组节点间连接的子集。 每个子集在STM神经形态网络的时间复用周期的多个时隙的不同时隙中形成。 结合在一起，节点间连接子集实现完全连接的神经网络。 突触时间复用神经元网络的方法包括提供神经织物并形成一组节间连接的子集。

公开(公告)号：US08577815B2

公开(公告)日：2013-11-05

申请号：US12604606

申请日：2009-10-23

申请人： Leandro G. Barajas ,  Youngkwan Cho ,  Narayan Srinivasa

发明人： Leandro G. Barajas ,  Youngkwan Cho ,  Narayan Srinivasa

IPC分类号： G06F15/18 ,  G06N3/00 ,  G06N3/12

CPC分类号： G06N3/02 ,  G06F17/18 ,  G06K9/00496 ,  G06K9/6251

摘要： A method and system for characterizing, detecting, and predicting or forecasting multiple target events from a past history of these events includes compressing temporal data streams into self-organizing map (SOM) clusters, and determining trajectories of the temporal streams via the clusters to predict the multiple target events. The system includes an evolutionary multi-objective optimization (EMO) module for processing the temporal data streams, which are obtained from a plurality of heterogeneous domains; a SOM module for characterizing the temporal data streams into self-organizing map clusters; and a target event prediction (TEP) module for generating prediction models of the map clusters. The SOM module employs a vector quantization method that places a set of vectors on a low-dimensional grid in an ordered fashion. The prediction models each include trajectories of the temporal data streams, and the system predicts the multiple target events using the trajectories.

摘要翻译： 用于从这些事件的过去历史表征，检测和预测或预测多个目标事件的方法和系统包括将时间数据流压缩为自组织映射（SOM）集群，以及通过集群确定时间流的轨迹以预测 多个目标事件。 该系统包括用于处理从多个异构域获得的时间数据流的进化多目标优化（EMO）模块; 用于将时间数据流表征为自组织映射簇的SOM模块; 以及用于生成地图簇的预测模型的目标事件预测（TEP）模块。 SOM模块采用矢量量化方法，其以有序的方式将一组向量放置在低维度网格上。 预测模型各自包括时间数据流的轨迹，并且系统使用轨迹来预测多个目标事件。

公开(公告)号：US08406989B1

公开(公告)日：2013-03-26

申请号：US12378393

申请日：2009-02-13

申请人： Rajan Bhattacharyya ,  Narayan Srinivasa

发明人： Rajan Bhattacharyya ,  Narayan Srinivasa

IPC分类号： G06F19/00 ,  G06T15/00

CPC分类号： B25J9/1666 ,  G05B2219/40317 ,  G05B2219/40367

摘要： A method for adaptive obstacle avoidance for articulated redundant robots is disclosed. The method comprises acts of calculating an obstacle avoidance vector for each of a set of limbs in a robot arm, and then applying the obstacle avoidance vector to constrain the inverse model in a robot controller. The obstacle avoidance vector incorporates factors including: (1) a distance and direction of each of a set of obstacles to the limb; and (2) when the limb is part of a kinematic chain of limbs, contributions from the obstacle avoidance vectors of all peripheral limbs in the kinematic chain. The method of the present invention was designed for use with the DIRECT model robot controller, but the method is generally applicable to any of a variety of robot controllers known in the art.

摘要翻译： 公开了一种用于铰接冗余机器人的自适应障碍物避免的方法。 该方法包括对机器人手臂中的一组肢体中的每一者计算障碍物回避矢量的动作，然后应用障碍物回避矢量来约束机器人控制器中的逆模型。 障碍物回避矢量包含以下因素：（1）肢体中一组障碍物的距离和方向; 和（2）当肢体是四肢运动链的一部分时，来自运动链中所有周围肢体的障碍物回避矢量的贡献。 本发明的方法被设计为与DIRECT模型机器人控制器一起使用，但是该方法通常适用于本领域已知的各种机器人控制器中的任何一种。

公开(公告)号：US08051018B1

公开(公告)日：2011-11-01

申请号：US11999427

申请日：2007-12-04

申请人： Narayan Srinivasa ,  Richard Stewart Ross

发明人： Narayan Srinivasa ,  Richard Stewart Ross

IPC分类号： G06F15/18 ,  G06N3/12 ,  G06N7/06

CPC分类号： G06N3/126

摘要： The present invention relates to a method for optimizing the design of a shape morphing structure using a genetic algorithm. The method includes defining design parameters of a surface having variable properties into a chromosome. The variable properties of the chromosome are the actual properties of the chromosome. The chromosome has a total of Nmax genes, where each gene corresponds to a variable property element in the surface. Additionally, each gene has n design parameters, wherein the design parameters are incremental changes to the actual properties of the chromosome. A genetic algorithm is employed to optimize the genes until a fitness level for at least one chromosome has been exceeded. When the fitness value for any chromosome in the population is above a predetermined threshold, then the design process is terminated and the final design solution[s] are the design parameters of the chromosome[s] that exceed the predetermined threshold value.

摘要翻译： 本发明涉及使用遗传算法优化形状变形结构的设计的方法。 该方法包括将具有可变特性的表面的设计参数定义到染色体中。 染色体的可变性质是染色体的实际性质。 染色体共有Nmax基因，其中每个基因对应于表面中的可变性质元素。 此外，每个基因具有n个设计参数，其中设计参数是对染色体的实际性质的增量变化。 使用遗传算法来优化基因，直到超过至少一个染色体的适应度水平为止。 当群体中任何染色体的适合度值高于预定阈值时，终止设计过程，并且最终设计解决方案是超过预定阈值的染色体的设计参数。

公开(公告)号：US07977906B1

公开(公告)日：2011-07-12

申请号：US12228579

申请日：2008-08-14

申请人： Narayan Srinivasa ,  Youngkwan Cho

发明人： Narayan Srinivasa ,  Youngkwan Cho

IPC分类号： G05B13/00

CPC分类号： G06N3/008 ,  G06K9/00664 ,  G06K9/00771

摘要： Described is a fault-tolerant electro-mechanical system that is able to saccade to a target by training and using a signal processing technique. The invention enables tracking systems, such as next generational cameras, to be developed for autonomous platforms and surveillance systems where environment conditions are unpredictable. The invention includes at least one sensor configured to relay a signal containing positional information of a stimulus. At least one actuator is configured to manipulate the sensor to enable the sensor to track the stimulus. A processing device is configured to receive positional information from each sensor and each actuator. The processing device sends a positional changing signal to at least one actuator and adjusts at least one positional changing signal according to the information from each sensor and each actuator to enable the actuator to cause the sensor to track the stimulus.

摘要翻译： 描述了能够通过训练和使用信号处理技术将目标扫视的容错机电系统。 本发明使得能够针对环境条件不可预测的自主平台和监视系统开发诸如下一代照相机的跟踪系统。 本发明包括至少一个传感器，其构造成中继包含刺激的位置信息的信号。 至少一个致动器被配置为操纵传感器以使传感器能够跟踪刺激。 处理装置被配置为从每个传感器和每个致动器接收位置信息。 处理装置向至少一个致动器发送位置改变信号，并且根据来自每个传感器和每个致动器的信息调整至少一个位置改变信号，以使致动器能够使传感器跟踪刺激。