Static model identification of quartz flexible accelerometer based on PSO⁃BP

隐含层节点数 / 个	训练集均方误差 / (m / s²)²
3	1.087 8 × 10^-5
4	4.252 1 × 10^-5
5	1.394 8 × 10^-5
6	2.066 1 × 10^-5
7	8.272 6 × 10^-5
8	8.351 7 × 10^-6
9	1.288 2 × 10^-5
10	6.133 6 × 10^-5
11	1.091 1 × 10^-4
12	2.112 4 × 10^-5

Table 2. Tab.2Static model coefficients identified by three algorithms

View table

View in Article

Table 2. Tab.2Static model coefficients identified by three algorithms

系数	LSM	GMDH	PSO⁃BP
$K_{0}$	2.274 5 × 10^-5	1.346 62 × 10^-5	1.306 17 × 10^-5
$K_{i}$	1	0.999 98	0.999 979
$K_{o}$		6.953 × 10^-5	6.869 24 × 10^-5
$K_{p}$		5.986 84 × 10^-6	5.320 93 × 10^-6
$K_{i o}$	2.532 2 × 10^-5	2.536 35 × 10^-5	2.584 44 × 10^-5
$K_{i p}$	5.911 5 × 10^-5	6.002 17 × 10^-5	6.090 3 × 10^-5
$K_{o p}$			7.145 91 × 10^-5
$K_{i i}$	4.825 4 × 10^-5	4.867 35 × 10^-5	4.817 09 × 10^-5
$K_{o o}$	3.653 9 × 10^-6	4.265 6 × 10^-6	3.411 39 × 10^-6
$K_{p p}$	3.583 9 × 10^-6	3.986 23 × 10^-6	4.561 5 × 10^-6
$K_{i i i}$	4.712 7 × 10^-6	4.786 86 × 10^-6	5.408 49 × 10^-6
$K_{o o o}$			4.022 96 × 10^-8
$K_{o i i}$			1.730 53 × 10^-7
$K_{p o o}$			6.581 32 × 10^-7
$K_{i p p}$			7.881 69 × 10^-7

Table 3. Tab.3Test set prediction error for BP and PSO⁃BP
View table
View in Article
Table 3. Tab.3Test set prediction error for BP and PSO⁃BP
算法 M_MAE / (m / s²) M_MSE / (m / s²)² R_RMSE/ (m / s²) M_MAPE / % $R^{2}$
BP 0.041 677 0.002 688 4 0.051 849 0.029 704 0.999 64
PSO⁃BP 0.002 536 4 9.178 × 10^-6 0.003 029 6 0.012 695 0.999 956

Tools

Get Citation

Copy Citation Text

Youzhi SHI, Renjian FENG, Xiaoting LI. Static model identification of quartz flexible accelerometer based on PSO⁃BP[J]. Metrology & Measurement Technology, 2025, 45(3): 78

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: Sensor Technology

Received: Mar. 3, 2025

Accepted: --

Published Online: Jul. 31, 2025

The Author Email:

DOI:10.11823/j.issn.1674-5795.2025.03.07

Topics

Table 1. Tab.1Mean square error calculation results of the training set with different number of hidden layer nodes

Table 1. Tab.1Mean square error calculation results of the training set with different number of hidden layer nodes

Table 2. Tab.2Static model coefficients identified by three algorithms

Table 2. Tab.2Static model coefficients identified by three algorithms

Table 3. Tab.3Test set prediction error for BP and PSO⁃BP

Table 3. Tab.3Test set prediction error for BP and PSO⁃BP