Nonlinear Averaging of Multi-Sensor Data

2003 ◽  
Author(s):  
Sugathevan Suranthiran ◽  
Suhada Jayasuriya
Keyword(s):  
Prior Knowledge ◽  
Sensor Fusion ◽  
Kalman Filtering ◽  
Sensor Data ◽  
Simulation Data ◽  
Model Based ◽  
Multiple Sensor ◽  
Multiple Measurements ◽  
Reliable Solution ◽  
Estimation Models

Considered in this paper is a framework for combining multiple sensor data to obtain a single inference. The task of fusing multi-sensor data is very challenging when no information about the sensor or estimation models is available. Kalman Filtering and other model-based techniques cannot be used to obtain a reliable inference. Linear Averaging of data is probably the simplest technique available, however, there is no guarantee that the fused measurement is, in fact, the best estimation. The problem will be worsened if one or more sensor measurements are faulty. In this paper, we analyze this problem and propose an effective multi-sensor fusion methodology. It is shown that a reliable solution can be obtained by nonlinearly averaging the multiple measurements. The proposed technique is well suited to identify outliers in the sensor measurements as well as to detect faulty sensor measurements. The developed algorithm is versatile in the sense that prior knowledge or information about sensors can be easily incorporated to improve the accuracy further. Illustrative examples and simulation data are presented to validate the proposed scheme.

scholarly journals Gas Detection and Identification Using Multimodal Artificial Intelligence Based Sensor Fusion

2021 ◽  
Vol 4 (1) ◽  
pp. 3
Author(s):  
Parag Narkhede ◽  
Rahee Walambe ◽  
Shruti Mandaokar ◽  
Pulkit Chandel ◽  
Ketan Kotecha ◽  
...  
Keyword(s):  
Sensor Fusion ◽  
Network Architecture ◽  
Cost Effective ◽  
Sensor Data ◽  
Dense Layer ◽  
Gaseous Emissions ◽  
Multiple Sensors ◽  
Single Output ◽  
Novel Approach ◽  
Single Sensor

With the rapid industrialization and technological advancements, innovative engineering technologies which are cost effective, faster and easier to implement are essential. One such area of concern is the rising number of accidents happening due to gas leaks at coal mines, chemical industries, home appliances etc. In this paper we propose a novel approach to detect and identify the gaseous emissions using the multimodal AI fusion techniques. Most of the gases and their fumes are colorless, odorless, and tasteless, thereby challenging our normal human senses. Sensing based on a single sensor may not be accurate, and sensor fusion is essential for robust and reliable detection in several real-world applications. We manually collected 6400 gas samples (1600 samples per class for four classes) using two specific sensors: the 7-semiconductor gas sensors array, and a thermal camera. The early fusion method of multimodal AI, is applied The network architecture consists of a feature extraction module for individual modality, which is then fused using a merged layer followed by a dense layer, which provides a single output for identifying the gas. We obtained the testing accuracy of 96% (for fused model) as opposed to individual model accuracies of 82% (based on Gas Sensor data using LSTM) and 93% (based on thermal images data using CNN model). Results demonstrate that the fusion of multiple sensors and modalities outperforms the outcome of a single sensor.

Low Cost Sensor Data Fusion in Omnidirectional Mobile Robot Feedback System to Improve the Navigation Accuracy

2014 ◽  
Vol 607 ◽  
pp. 791-794 ◽  
Author(s):  
Wei Kang Tey ◽  
Che Fai Yeong ◽  
Yip Loon Seow ◽  
Eileen Lee Ming Su ◽  
Swee Ho Tang
Keyword(s):  
Mobile Robot ◽  
Sensor Fusion ◽  
Low Cost ◽  
Feedback System ◽  
Sensor Data ◽  
Research Setting ◽  
Fusion Technique ◽  
Omnidirectional Mobile Robot ◽  
The Cost ◽  
Navigation Accuracy

Omnidirectional mobile robot has gained popularity among researchers. However, omnidirectional mobile robot is rarely been applied in industry field especially in the factory which is relatively more dynamic than normal research setting condition. Hence, it is very important to have a stable yet reliable feedback system to allow a more efficient and better performance controller on the robot. In order to ensure the reliability of the robot, many of the researchers use high cost solution in the feedback of the robot. For example, there are researchers use global camera as feedback. This solution has increases the cost of the robot setup fee to a relatively high amount. The setup system is also hard to modify and lack of flexibility. In this paper, a novel sensor fusion technique is proposed and the result is discussed.

Routine Clustering of Mobile Sensor Data Facilitates Psychotic Relapse Prediction in Schizophrenia Patients (Preprint)

2021 ◽  
Author(s):  
Joanne Zhou ◽  
Bishal Lamichhane ◽  
Dror Ben-Zeev ◽  
Andrew Campbell ◽  
Akane Sano
Keyword(s):  
Prediction Models ◽  
Gaussian Mixture ◽  
Mobile Sensing ◽  
Cluster Models ◽  
Sensor Data ◽  
Sensing Data ◽  
Model Based ◽  
Clustering Model ◽  
Relapse Prediction ◽  
Lower Variability

BACKGROUND Behavioral representations obtained from mobile sensing data could be helpful for the prediction of an oncoming psychotic relapse in schizophrenia patients and delivery of timely interventions to mitigate such relapse. OBJECTIVE In this work, we aim to develop clustering models to obtain behavioral representations from continuous multimodal mobile sensing data towards relapse prediction tasks. The identified clusters could represent different routine behavioral trends related to daily living of patients as well as atypical behavioral trends associated with impending relapse. METHODS We used the mobile sensing data obtained in the CrossCheck project for our analysis. Continuous data from six different mobile sensing-based modalities (e.g. ambient light, sound/conversation, acceleration etc.) obtained from a total of 63 schizophrenia patients, each monitored for up to a year, were used for the clustering models and relapse prediction evaluation. Two clustering models, Gaussian Mixture Model (GMM) and Partition Around Medoids (PAM), were used to obtain behavioral representations from the mobile sensing data. These models have different notions of similarity between behaviors as represented by the mobile sensing data and thus provide differing behavioral characterizations. The features obtained from the clustering models were used to train and evaluate a personalized relapse prediction model using Balanced Random Forest. The personalization was done by identifying optimal features for a given patient based on a personalization subset consisting of other patients who are of similar age. RESULTS The clusters identified using the GMM and PAM models were found to represent different behavioral patterns (such as clusters representing sedentary days, active but with low communications days, etc.). While GMM based models better characterized routine behaviors by discovering dense clusters with low cluster spread, some other identified clusters had a larger cluster spread likely indicating heterogeneous behavioral characterizations. PAM model based clusters on the other hand had lower variability of cluster spread, indicating more homogeneous behavioral characterization in the obtained clusters. Significant changes near the relapse periods were seen in the obtained behavioral representation features from the clustering models. The clustering model based features, together with other features characterizing the mobile sensing data, resulted in an F2 score of 0.24 for the relapse prediction task in a leave-one-patient-out evaluation setting. This obtained F2 score is significantly higher than a random classification baseline with an average F2 score of 0.042. CONCLUSIONS Mobile sensing can capture behavioral trends using different sensing modalities. Clustering of the daily mobile sensing data may help discover routine as well as atypical behavioral trends. In this work, we used GMM and PAM-based cluster models to obtain behavioral trends in schizophrenia patients. The features derived from the cluster models were found to be predictive for detecting an oncoming psychotic relapse. Such relapse prediction models can be helpful to enable timely interventions.

scholarly journals Sensor Fusion for Recognition of Activities of Daily Living

Sensors ◽  
2018 ◽  
Vol 18 (11) ◽  
pp. 4029 ◽  
Author(s):  
Jiaxuan Wu ◽  
Yunfei Feng ◽  
Peng Sun
Keyword(s):  
Time Series ◽  
Sensor Fusion ◽  
Daily Living ◽  
Single Point ◽  
Recognition System ◽  
Sensor Data ◽  
Multiple Sensors ◽  
Life Patterns ◽  
Functional Capabilities ◽  
Correction Technique

Activity of daily living (ADL) is a significant predictor of the independence and functional capabilities of an individual. Measurements of ADLs help to indicate one’s health status and capabilities of quality living. Recently, the most common ways to capture ADL data are far from automation, including a costly 24/7 observation by a designated caregiver, self-reporting by the user laboriously, or filling out a written ADL survey. Fortunately, ubiquitous sensors exist in our surroundings and on electronic devices in the Internet of Things (IoT) era. We proposed the ADL Recognition System that utilizes the sensor data from a single point of contact, such as smartphones, and conducts time-series sensor fusion processing. Raw data is collected from the ADL Recorder App constantly running on a user’s smartphone with multiple embedded sensors, including the microphone, Wi-Fi scan module, heading orientation of the device, light proximity, step detector, accelerometer, gyroscope, magnetometer, etc. Key technologies in this research cover audio processing, Wi-Fi indoor positioning, proximity sensing localization, and time-series sensor data fusion. By merging the information of multiple sensors, with a time-series error correction technique, the ADL Recognition System is able to accurately profile a person’s ADLs and discover his life patterns. This paper is particularly concerned with the care for the older adults who live independently.

scholarly journals Three Revised Kalman Filtering Models for Short-Term Rail Transit Passenger Flow Prediction

2016 ◽  
Vol 2016 ◽  
pp. 1-10 ◽  
Author(s):  
Pengpeng Jiao ◽  
Ruimin Li ◽  
Tuo Sun ◽  
Zenghao Hou ◽  
Amir Ibrahim
Keyword(s):  
Real Time ◽  
Nonparametric Regression ◽  
Kalman Filtering ◽  
Rail Transit ◽  
Correction Coefficient ◽  
Short Term ◽  
Transit System ◽  
Filtering Method ◽  
Passenger Flow ◽  
Model Based

Short-term prediction of passenger flow is very important for the operation and management of a rail transit system. Based on the traditional Kalman filtering method, this paper puts forward three revised models for real-time passenger flow forecasting. First, the paper introduces the historical prediction error into the measurement equation and formulates a revised Kalman filtering model based on error correction coefficient (KF-ECC). Second, this paper employs the deviation between real-time passenger flow and corresponding historical data as state variable and presents a revised Kalman filtering model based on Historical Deviation (KF-HD). Third, the paper integrates nonparametric regression forecast into the traditional Kalman filtering method using a Bayesian combined technique and puts forward a revised Kalman filtering model based on Bayesian combination and nonparametric regression (KF-BCNR). A case study is implemented using statistical passenger flow data of rail transit line 13 in Beijing during a one-month period. The reported prediction results show that KF-ECC improves the applicability to historical trend, KF-HD achieves excellent accuracy and stability, and KF-BCNR yields the best performances. Comparisons among different periods further indicate that results during peak periods outperform those during nonpeak periods. All three revised models are accurate and stable enough for on-line predictions, especially during the peak periods.

A Toolkit to Assist Researchers to More Efficiently Conduct Experiments Assessing Human State

2017 ◽  
Vol 61 (1) ◽  
pp. 2032-2036 ◽  
Author(s):  
Bethany K. Bracken ◽  
Noa Palmon ◽  
David Koelle ◽  
Mike Farry
Keyword(s):  
Real Time ◽  
Team Performance ◽  
Team Member ◽  
Sensor Data ◽  
Team Members ◽  
Augmentation Strategies ◽  
Experimental Protocols ◽  
Multiple Sensor ◽  
Support Research

For teams to perform effectively, individuals must focus on their own tasks, while simultaneously maintaining awareness of other team members. Researchers studying and attempting to optimize performance of teams as well as individual team members use assessments of behavioral, neurophysiological, and physiological signals that correlate with individual and team performance. However, synchronizing data from multiple sensor devices can be difficult, and building and using models to assess human states of interest can be time-consuming and non-intuitive. To assist researchers, we built an Adaptable Toolkit for the Assessment and Augmentation of Performance by Teams in Real Time (ADAPTER), which provides a framework that flexibly integrates sensors and fuses sensor data to assess performance. ADAPTER flexibly integrates current and emerging sensors; assists researchers in creating and implementing models that support research on performance and the development of augmentation strategies; and enables comprehensive and holistic characterization of team member performance during real-time experimental protocols.

scholarly journals A Hybrid Approach: Dynamic Diagnostic Rules for Sensor Systems in Industry 4.0 Generated by Online Hyperparameter Tuned Random Forest

2020 ◽  
Author(s):  
Ahlam Mallak ◽  
Madjid Fathi
Keyword(s):  
Random Forest ◽  
Random Search ◽  
Hybrid Approach ◽  
Fault Detection And Diagnosis ◽  
Data Driven ◽  
Sensor Data ◽  
Sensor Systems ◽  
Hydraulic Test ◽  
Model Based ◽  
Detection And Diagnosis

In this work, A hybrid component Fault Detection and Diagnosis (FDD) approach for industrial sensor systems is established and analyzed, to provide a hybrid schema that combines the advantages and eliminates the drawbacks of both model-based and data-driven methods of diagnosis. Moreover, spotting the light on a new utilization of Random Forest (RF) together with model-based diagnosis, beyond its ordinary data-driven application. RF is trained and hyperparameter tuned using 3-fold cross-validation over a random grid of parameters using random search, to finally generate diagnostic graphs as the dynamic, data-driven part of this system. Followed by translating those graphs into model-based rules in the form of if-else statements, SQL queries or semantic queries such as SPARQL, in order to feed the dynamic rules into a structured model essential for further diagnosis. The RF hyperparameters are consistently updated online using the newly generated sensor data, in order to maintain the dynamicity and accuracy of the generated graphs and rules thereafter. The architecture of the proposed method is demonstrated in a comprehensive manner, as well as the dynamic rules extraction phase is applied using a case study on condition monitoring of a hydraulic test rig using time series multivariate sensor readings.

Sign in / Sign up

Export Citation Format

Share Document