Tata Consultancy Services Patente

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Load disaggregation is useful for both the consumers and producers of energy. The present-day supervised learning models for load disaggregation necessitate the learning of models for every appliance load of interest, which incurs high computational costs. Embodiments of the present disclosure implement a Restricted Boltzmann Machine (RBM) based source-separation model with application to load disaggregation of appliances of interest. Representations of appliance of interest are learnt, between the power aggregate data and the appliance signatures, to output the mapping of data representations on the appliance signatures, for load disaggregation. Discriminative ability for each load/appliance of interest is achieved by adding the free energies of softmax layers of the RBM on other loads/appliance, as a discriminating gradient to the approximate gradients obtained on the load under consideration.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

E-commerce industry is currently expanding rapidly, worldwide. A process of generating product copy for grocery items, which is very challenging as food items, do not have features in common, unlike fashion products. A data associated with one or more grocery products is received as an input. The data is processed to obtain one or more sorted similar grocery products. One or more relevant attributes and an allergen information associated with the one or more sorted similar grocery products are extracted. A vocabulary model is created based the one or more relevant attributes and the allergen information associated with the one or more sorted similar grocery products. The vocabulary model is validated based on one or more assigned weights on a training data. The one or more descriptive copies associated with grocery products are generated by mapping the validated vocabulary model with the training data.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Continuous monitoring of subject's cardiac system using biological signal(s) (BS) during day-to-day activities is essential for managing personal cardiac health/disorders, etc. Conventional systems/methods lack in improvising overall classification results and configured for specific device/signal say ECG or PPG and so on. Present disclosure provides systems and methods for classifying BS obtained from users, wherein BS are preprocessed to obtain filtered signals (FS). Corresponding feature extraction module is utilized for feature set extraction based on features in FS. The feature set is reduced and segmented into test and training data. Biological signal classification model(s) are generated using training data and a BCM is applied on test data to classify biological signals (BS) as one of Atrial Fibrillation (AF), a non-AF, a cardiac arrythmia disorder, or ischemia. Accelerometer features of connected device associated with the users can be obtained to detect activities which in conjunction with the BCM's output improvises above classification.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

In industrial inspection scenarios, early detection of machine malfunction is extremely essential as it helps in preventing any significant damage and the associated economic losses. Embodiments herein provide a method and system for an acoustic based anomaly detection in industrial machines using a beamforming and a sequential transform learning. Herein, the system employs two-stage multichannel source separation technique that uses the well-known delay and sum beamforming followed by a recent data-driven sequential transform learning (STL) approach to obtain clean sources. The STL is a solution to linear state-space model where operators/matrices are learnt from data and is used here to model the dynamics of time-varying source signals for source separation. Subsequently, a reference template matching is employed on each separated source to detect an anomaly. The numerical results obtained with the Malfunctioning Industrial Machine Investigation and Inspection (MIMII) dataset demonstrate superior performance for source separation and anomaly detection.

This disclosure relates generally to system and method for identifying suitable equipment for pelletizations of particles. Conventional approaches for identifying pelletization equipment for a process are cumbersome and unmethodical. Even though, methods of identifying optimal operating conditions in a particular equipment for a desired product specification are widely discussed in literature, these research works lack in considering practical criteria like equipment cost, equipment capacity, equipment maintenance, etc., in addition to product specifications while deciding a suitable pelletization equipment. Present disclosure provides system and method that involve optimization to identify the best range of operating parameters in a pelletization equipment to achieve desired product size distribution for a given feed size distribution. The method is then used in a system for comparing different pelletization equipment based on different criteria and identifying a suitable pelletization equipment amongst them for use in obtaining pellets with desired product attributes.

This disclosure relates to a method and system for hybrid data augmentation for estimating performance degradation in industrial plant. Performance degradation in industrial plants cannot be measured by sensors or laboratory measurements and there are no methods to annotate performance degradation state. The embodiments of the present disclosure provide a knowledge-based data augmentation that use physics based information to model performance degradation. The disclosed method augments high fidelity data with knowledge-based methods into high and low confidence data which are used to calculate performance score of high confidence data. A physics-informed machine learning model is trained on high confidence data. The resulting model is then used to predict performance score for low confidence data. The model is further used for training prognostics and diagnostics models to predict and identify root causes responsible for performance degradation. The disclosed method is used for predicting ring formation inside the sponge iron kiln.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

This disclosure relates to application based design of novel materials. Conventional methods utilize laborious experimentation or costly first principles calculations. Conventional data driven techniques use point cloud-based representation for crystal structures, that suffers from permutation variance which is not inbuilt in a material's representation, the DL model has to learn invariance which may be inaccurate. Other methods use image based representation for crystal structures and separate images for each element type to represent the basis, which is memory and time intensive. Since each element is represented by its own image, it is difficult for model to learn chemical environment and neighborhood pattern of each element. The embodiments used image based representation of materials consistent with physical principles. Also, embodiments utilize elements matrix to obtain atoms and their positions from basis images. Thus, any material, irrespective of lattice geometry, and number and types of elements, is represented by only two images.

Zusammenfassung wird geladen …

This disclosure relates generally to in-silico modeling of hemodynamic patterns of physiologic blood flow. Conventional cardiovascular hemodynamic models depend on neuromodulation schemes (baroreflex autoregulation) and threshold parameters of neuromodulation correlate with physical activities. Thus these models may not work practically for a large set of people due to dependency on prior knowledge of these parameters. The present disclosure enables estimating blood pressure of a subject by estimating cardiac parameters based on the morphology of ECG signal associated with the subject and hence activation delays in cardiac chambers of the in-silico model is reproduced purposefully. In accordance with the present disclosure, the blood pressure of the subject can be estimated using only the ECG signal even if the signal is missed for some time instance(s) or is noisy.

Zusammenfassung wird geladen …

The embodiments of present disclosure address a need for a framework to holistically utilize storage capacity of an Energy Storage System (ESS) to serve forecast errors of several Renewable Energy Generators (REGens) participating in a day-ahead market. Embodiments herein provide a method and system for optimizing the operation and price of an Energy Storage as a Service (ESaaS) framework. In anticipation of the forecast errors from REGens, the ESS operator takes suitable countermeasures such as charging/discharging of storage system through market transactions. This is done in a way to reduce imbalance in the market commitments made by individual REGens without reserving any storage volume for each REGen. Further, the system is configured to schedule the storage, determine the settlement volumes, and decide the service prices. The disclosed ESaaS framework is beneficial for all entities such as REGens (revenue outflow decreases), system operator (imbalance volume reduces), and ESS (revenue earned increases).

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

A growing need for inferencing to be run on fog devices exists, in order to reduce the upstream network traffic. However, being computationally constrained in nature, executing complex deep inferencing models on such devices has been proved difficult. A system and method for partitioning of deep convolution neural network for execution of computationally constraint devices at a network edge has been provided. The system is configured to use depth wise input partitioning of convolutional operations in deep convolutional neural network (DCNN). The convolution operation is performed based on an input filter depth and number of filters for determining the appropriate parameters for partitioning based on an inference speedup method. The system uses a master-slave network for partitioning the input. The system is configured to address these problems by depth wise partitioning of input which ensures speedup inference of convolution operations by reducing pixel overlaps.

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …

Zusammenfassung wird geladen …