Abstract:Micro-electro-mechanical system inertial measurement unit (MEMS-IMU), a core component in many navigation systems, directly determines the accuracy of inertial navigation system; however, MEMS-IMU system is often affected by various factors such as environmental noise, electronic noise, mechanical noise and manufacturing error. These can seriously affect the application of MEMS-IMU used in different fields. Focus has been on MEMS gyro since it is an essential and, yet, complex sensor in MEMS-IMU which is very … Show more
“…For real-time online signal processing, the results of the proposed scheme are compared with small-scale networks, and the use of MLP network and leaky ReLU activation function with simple calculations are sufficient to achieve the corresponding error suppression effect. From Table 2, when smallscale recurrent neural networks are used [23,24], the error suppression effect obtained is similar to that of the MLP network used in this article, and with large-scale networks [22,26], the error suppression effect obtained is better than that of the proposed method. The result of the comparison shows that the compensation effect of the adopted scheme at the algorithm level is effective and reasonable.…”
Section: Implementation Details and Experimental Resultssupporting
confidence: 59%
“…The computational complexity of the error compensation scheme based on neural networks is large, which is a challenge for real-time online applications. Compared with related works [22][23][24][25][26], the network model and activation function used in this paper are simpler and less complex. When the number of neurons in the input layer and the number of neurons in the hidden layer are N and H, the shape of the input vector is (1, N), the shape of the hidden layer parameter matrix is (N, H), and the shape of the hidden layer feature matrix is (1, N).…”
Section: Circuit-level Realization and Analysis Of The Error Compensation Schemementioning
confidence: 99%
“…This type of error directly affects the stability of the output signals and is difficult to be processed directly through device calibration [8]. Therefore, the modeling and compensation schemes of the nonlinear error components are widely studied and two mainstream research schemes [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26] are formed, namely, (1) establishing a statistical model and performing error compensating and (2) error compensation schemes based on machine learning or deep learning.…”
Section: Introductionmentioning
confidence: 99%
“…The accuracy improvement of statistical methods requires accurate analysis and modeling of error components, while another research idea is to obtain model structure and model parameters with machine learning schemes. The researchers use multilayer perception machines [20,21], recurrent neural networks [22][23][24], mixed deep learning networks [25,26], etc. to establish error models.…”
Section: Introductionmentioning
confidence: 99%
“…The nonlinear error in the composition of measurement error is discussed. In related research studies [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26], the nonlinear error is modeled as a time series model, in which the error at the current moment is related to the observation values at previous moments. The error compensation scheme discussed in this paper is also based on the analysis of the time series model.…”
Section: Problem Description and Introduction Of Error Compensation Schemementioning
Nonlinear errors of sensor output signals are common in the field of inertial measurement and can be compensated with statistical models or machine learning models. Machine learning solutions with large computational complexity are generally offline or implemented on additional hardware platforms, which are difficult to meet the high integration requirements of microelectromechanical system inertial sensors. This paper explored the feasibility of an online compensation scheme based on neural networks. In the designed solution, a simplified small-scale network is used for modeling, and the peak-to-peak value and standard deviation of the error after compensation are reduced to 17.00% and 16.95%, respectively. Additionally, a compensation circuit is designed based on the simplified modeling scheme. The results show that the circuit compensation effect is consistent with the results of the algorithm experiment. Under SMIC 180 nm complementary metal-oxide semiconductor (CMOS) technology, the circuit has a maximum operating frequency of 96 MHz and an area of 0.19 mm2. When the sampling signal frequency is 800 kHz, the power consumption is only 1.12 mW. This circuit can be used as a component of the measurement and control system on chip (SoC), which meets real-time application scenarios with low power consumption requirements.
“…For real-time online signal processing, the results of the proposed scheme are compared with small-scale networks, and the use of MLP network and leaky ReLU activation function with simple calculations are sufficient to achieve the corresponding error suppression effect. From Table 2, when smallscale recurrent neural networks are used [23,24], the error suppression effect obtained is similar to that of the MLP network used in this article, and with large-scale networks [22,26], the error suppression effect obtained is better than that of the proposed method. The result of the comparison shows that the compensation effect of the adopted scheme at the algorithm level is effective and reasonable.…”
Section: Implementation Details and Experimental Resultssupporting
confidence: 59%
“…The computational complexity of the error compensation scheme based on neural networks is large, which is a challenge for real-time online applications. Compared with related works [22][23][24][25][26], the network model and activation function used in this paper are simpler and less complex. When the number of neurons in the input layer and the number of neurons in the hidden layer are N and H, the shape of the input vector is (1, N), the shape of the hidden layer parameter matrix is (N, H), and the shape of the hidden layer feature matrix is (1, N).…”
Section: Circuit-level Realization and Analysis Of The Error Compensation Schemementioning
confidence: 99%
“…This type of error directly affects the stability of the output signals and is difficult to be processed directly through device calibration [8]. Therefore, the modeling and compensation schemes of the nonlinear error components are widely studied and two mainstream research schemes [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26] are formed, namely, (1) establishing a statistical model and performing error compensating and (2) error compensation schemes based on machine learning or deep learning.…”
Section: Introductionmentioning
confidence: 99%
“…The accuracy improvement of statistical methods requires accurate analysis and modeling of error components, while another research idea is to obtain model structure and model parameters with machine learning schemes. The researchers use multilayer perception machines [20,21], recurrent neural networks [22][23][24], mixed deep learning networks [25,26], etc. to establish error models.…”
Section: Introductionmentioning
confidence: 99%
“…The nonlinear error in the composition of measurement error is discussed. In related research studies [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26], the nonlinear error is modeled as a time series model, in which the error at the current moment is related to the observation values at previous moments. The error compensation scheme discussed in this paper is also based on the analysis of the time series model.…”
Section: Problem Description and Introduction Of Error Compensation Schemementioning
Nonlinear errors of sensor output signals are common in the field of inertial measurement and can be compensated with statistical models or machine learning models. Machine learning solutions with large computational complexity are generally offline or implemented on additional hardware platforms, which are difficult to meet the high integration requirements of microelectromechanical system inertial sensors. This paper explored the feasibility of an online compensation scheme based on neural networks. In the designed solution, a simplified small-scale network is used for modeling, and the peak-to-peak value and standard deviation of the error after compensation are reduced to 17.00% and 16.95%, respectively. Additionally, a compensation circuit is designed based on the simplified modeling scheme. The results show that the circuit compensation effect is consistent with the results of the algorithm experiment. Under SMIC 180 nm complementary metal-oxide semiconductor (CMOS) technology, the circuit has a maximum operating frequency of 96 MHz and an area of 0.19 mm2. When the sampling signal frequency is 800 kHz, the power consumption is only 1.12 mW. This circuit can be used as a component of the measurement and control system on chip (SoC), which meets real-time application scenarios with low power consumption requirements.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.