A MEMS IMU De-Noising Method Using Long Short Term Memory Recurrent Neural Networks (LSTM-RNN)

Jiang, Changhui; Shuai, Chen; Chen, Yuwei; Zhang, Boya; Feng, Ziyi; Zhou, Hui; Bo, Yuming

doi:10.3390/s18103470

Cited by 93 publications

(74 citation statements)

References 38 publications

(84 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Most of the deep learning studies focus on the imagebased applications such as remote sensing, image matching and classifications, and detection of road features by combining different sources [16][17][18][19]. The long short-term memory (LSTM) methods and a variance of recurrent neural networks (RNN) were mostly applied for the navigation of ground vehicles, UAVs, and robotics [20][21][22]. This is because the navigation data is provided in time series, which is suitable for the LSTM model.…”

Section: Framework Of Deep Learningmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning-Based GNSS Network-Based Real-Time Kinematic Improvement for Autonomous Ground Vehicle Navigation

Kim

Bae

2019

Journal of Sensors

View full text Add to dashboard Cite

Much navigation over the last several decades has been aided by the global navigation satellite system (GNSS). In addition, with the advent of the multi-GNSS era, more and more satellites are available for navigation purposes. However, the navigation is generally carried out by point positioning based on the pseudoranges. The real-time kinematic (RTK) and the advanced technology, namely, the network RTK (NRTK), were introduced for better positioning and navigation. Further improved navigation was also investigated by combining other sensors such as the inertial measurement unit (IMU). On the other hand, a deep learning technique has been recently evolving in many fields, including automatic navigation of the vehicles. This is because deep learning combines various sensors without complicated analytical modeling of each individual sensor. In this study, we structured the multilayer recurrent neural networks (RNN) to improve the accuracy and the stability of the GNSS absolute solutions for the autonomous vehicle navigation. Specifically, the long short-term memory (LSTM) is an especially useful algorithm for time series data such as navigation with moderate speed of platforms. From an experiment conducted in a testing area, the LSTM algorithm developed the positioning accuracy by about 40% compared to GNSS-only navigation without any external bias information. Once the bias is taken care of, the accuracy will significantly be improved up to 8 times better than the GNSS absolute positioning results. The bias terms of the solution need to be estimated within the model by optimizing the layers as well as the nodes each layer, which should be done in further research.

show abstract

Section: Framework Of Deep Learningmentioning

confidence: 99%

“…A recurrent neural network (RNN) is an algorithm that is composed of a circulation structure [20,22,23,27]. The hidden nodes have directional information; thus, past and current data are connected to constitute a circulation.…”

Section: Framework Of Deep Learningmentioning

confidence: 99%

Deep Learning-Based GNSS Network-Based Real-Time Kinematic Improvement for Autonomous Ground Vehicle Navigation

Kim

Bae

2019

Journal of Sensors

View full text Add to dashboard Cite

show abstract

“…However, for a standalone GPS receiver, signal challenging environment might hinder its extensive applications, for instance, NLOS (none-ofsight) and multipath (MP) signal in urban canyons and dynamic stress under high dynamic; these negative factors may affect the signal availability or navigation solution accuracy and integrity. Without enough satellites with "clean" signals available, the receiver will fail to output correct PNT information [2][3][4][5][6][7][8]. For instance, the NLOS signals will cause the pseudorange bias, signal blockage will influence the satellite geometry distribution, and the MP will also influence the errors in pseudorange measurements [2][3][4][5][6][7][8][9][10][11][12][13][14].…”

Section: Introductionmentioning

confidence: 99%

“…e SINS processes the angular and acceleration measurements from the inertial measurement unit (IMU), and then 3D georeference information is generated without transmitting or receiving any outside signals [6][7][8]. However, the random noises contained in the angular and acceleration measurements will lead to the navigation solution errors accumulating dramatically over time [6][7][8]. erefore, the SINS is usually integrated with the GPS for providing more stable and reliable navigation solutions.…”

Section: Introductionmentioning

confidence: 99%

A New Efficient Filtering Model for GPS/SINS Ultratight Integration System

Wang

Jiang

2020

Mathematical Problems in Engineering

Self Cite

View full text Add to dashboard Cite

Global Positioning System (GPS) and strap-down inertial navigation system (SINS) are recognized as highly complementary and widely employed in the community. The GPS has the advantage of providing precise navigation solutions without divergence, but the GPS signals might be blocked and attenuated. The SINS is a totally self-contained navigation system which is hardly disturbed. The GPS/SINS integration system could utilize the advantages of both the GPS and SINS and provide more reliable navigation solutions. According to the data fusion strategies, the GPS/SINS integrated system could be divided into three different modes: loose, tight, and ultratight integration (LI, TI, and UTC). In the loose integration mode, position and velocity difference from the GPS and SINS are employed to compose measurement vector, in which the vector dimension has nothing to do with the amount of the available satellites. However, in the tight and ultratight modes, difference of pseudoranges and pseudorange rates from the GPS and SINS are employed to compose the measurement vector, in which the measurement vector dimension increases with the amount of available satellites. In addition, compared with the loose integration mode, clock bias and drift are included in the integration state model. The two characteristics magnify the computation load of the tight and ultratight modes. In this paper, a new efficient filter model was proposed and evaluated. Two schemes were included in this design for reducing the computation load. Firstly, a difference between pseudorange measurements was determined, by which clock bias and drift were excluded from the integration state model. This step reduced the dimension of the state vector. Secondly, the integration filter was divided into two subfilters: pseudorange subfilter and pseudorange rate subfilter. A federated filter was utilized to estimate the state errors optimally. In the second step, the two subfilters could run in parallel and the measurement vector was divided into two subvectors with lower dimension. A simulation implemented in MATLAB software was conducted to evaluate the performance of the new efficient integration method in UTC. The simulation results showed that the method could reduce the computation load with the navigation solutions almost unchanged.

show abstract

“…Specially, Long Short-Term Memory Recurrent Neural Network (LSTM-RNN) and Simple Reduced Unit Recurrent Neural Network (SRU-RNN) were employed in MEMS gyroscope noise suppressing. Both LSTM-RNN and SRU-RNN were popular variants of RNN, and they both performed better than conventional machine learning or regression method [32,33]. However, it is interesting to explore a more feasible RNN structure in this application.…”

Section: Introductionmentioning

confidence: 99%

A MEMS Gyroscope Noise Suppressing Method Using Neural Architecture Search Neural Network

Zhu

Jiang

2019

Mathematical Problems in Engineering

Self Cite

View full text Add to dashboard Cite

Inertial measurement unit (IMU) (an IMU usually contains three gyroscopes and accelerometers) is the key sensor to construct a selfcontained inertial navigation system (INS). IMU manufactured through the Micromechanics Electronics Manufacturing System (MEMS) technology becomes more popular, due to its smaller column, lower cost, and gradually improved accuracy. However, limited by the manufacturing technology, the MEMS IMU raw measurement signals experience complicated noises, which cause the INS navigation solution errors diverge dramatically over time. For addressing this problem, an advanced Neural Architecture Search Recurrent Neural Network (NAS-RNN) was employed in the MEMS gyroscope noise suppressing. NAS-RNN was the recently invented artificial intelligence method for time series problems in data science community. Different from conventional method, NAS-RNN was able to search a more feasible architecture for selected application. In this paper, a popular MEMS IMU STIM300 was employed in the testing experiment, and the sampling frequency was 125 Hz. e experiment results showed that the NAS-RNN was effective for MEMS gyroscope denoising; the standard deviation values of denoised three-axis gyroscope measurements decreased by 44.0%, 34.1%, and 39.3%, respectively. Compared with the Long Short-Term Memory Recurrent Neural Network (LSTM-RNN), the NAS-RNN obtained further decreases by 28.6%, 3.7%, and 8.8% in standard deviation (STD) values of the signals. In addition, the attitude errors decreased by 26.5%, 20.8%, and 16.4% while substituting the LSTM-RNN with the NAS-RNN.

show abstract

A MEMS IMU De-Noising Method Using Long Short Term Memory Recurrent Neural Networks (LSTM-RNN)

Cited by 93 publications

References 38 publications

Deep Learning-Based GNSS Network-Based Real-Time Kinematic Improvement for Autonomous Ground Vehicle Navigation

Deep Learning-Based GNSS Network-Based Real-Time Kinematic Improvement for Autonomous Ground Vehicle Navigation

A New Efficient Filtering Model for GPS/SINS Ultratight Integration System

A MEMS Gyroscope Noise Suppressing Method Using Neural Architecture Search Neural Network

Contact Info

Product

Resources

About