基于PSO-RNN算法的多级感应线圈炮非参数建模与出口速度预测

doi:10.12382/bgxb.2024.0616

摘要/Abstract

摘要： 针对多级同步感应线圈发射器建模涉及多物理场耦合、现有优化方法迭代时间长等问题,基于粒子群优化-循环神经网络(Particle Swarm Optimization-Recurrent Neural Network,PSO-RNN)算法建立多级同步感应线圈发射器非参数模型,并进行电枢出口速度预测。通过正交结合随机实验的方法,获得以线圈匝数、触发时间、触发位置为输入,出口速度为输出的样本集;采用循环神经网络算法对样本集进行训练并建立非参数模型;通过粒子群优化算法进一步优化RNN神经网络参数,提高非参数模型的预测性能;采用建立的模型预测出口速度并与实验结果对比。结果表明:所建立非参数模型的均方预测误差、平均绝对百分比误差、均方根误差分别为0.0028、0.036、2.18,且经过PSO优化后模型的3项评价指标分别降低39%、38%、46%,提高了预测性能;PSO-RNN非参数模型的一致性较好且预测的平均值与实验测得的出口速度相差1.2m/s,误差百分比为1.8%,小于标准值5%。将PSO-RNN算法用于同步感应线圈发射器的非参数建模可行且对出口速度的预测较为准确,可为多级同步感应线圈发射器的工程设计提供新思路。

关键词: 多级同步感应线圈炮, 非参数模型, 循环神经网络, 粒子群优化, 出口速度预测

Abstract:

An non-parametric model of multi-stage synchronous induction coilgun (MSSICG) based on the particle swarm optimization and recurrent neural network (PSO-RNN) algorithm is proposed to solve the problems such as multi-physics field coupling and long iteration time of existing optimization methods.And the ejection velocity of the armature is also predicted by the model.A sample set with the turns per coil,triggering time and trigger position as inputs and the ejection velocity as output is obtained through the orthogonal and random experiments.The RNN algorithm is used to train the sample set and the non-parametric model is established.The parameters of the RNN model are further optimized by the PSO algorithm,to improve the prediction performance of the non-parametric model.The ejection velocity of the armature is predicted using the proposed PSO-RNN model and compared with the experimental result.The MSPE,MAPE,and RMSE of the non-parametric model are 0.0028,0.036,and 2.18,respectively,which are reduced by 39%,38%,and 46% after the optimization of PSO.The difference between the predicted and experimental velocities is 1.2m/s with the error percentage of 1.8%,which is less than 5%.The study provides a novel idea for the modelling and engineering design of MSSICG.

Key words: multi-stage synchronous induction coilgun, non-parametric model, recurrent neural network, particle swarm optimization, muzzle velocity

中图分类号:

TM153

秦涛涛, 季思源, 雷琳, 郑占锋. 基于PSO-RNN算法的多级感应线圈炮非参数建模与出口速度预测[J]. 兵工学报, 2025, 46(7): 240616-.

QIN Taotao, JI Siyuan, LEI Lin, ZHENG Zhanfeng. Non-parametric Modelling and Muzzle Velocity Prediction of Multi-stage Induction Coilgun based on PSO-RNN Algorithm[J]. Acta Armamentarii, 2025, 46(7): 240616-.

图/表 23

参考文献 24

[1]	聂世雄. 集成式液压平衡电磁发射装置的非线性建模及控制[J]. 电工技术学报, 2023, 38(4):852-864.
	NIE S X. Nonlinear modelling and control of an integrated hydraulically balanced electromagnetic transmitter[J]. Journal of Electrotechnology, 2023, 38(4):852-864. (in Chinese)
[2]	RAM R, THOMAS M J.Computational studies on an induction coilgun[C]//Proceedings of the 2021 IEEE Pulsed Power Conference (PPC).Denver,CO,US:IEEE,2021:1-4.
[3]	马伟明, 鲁军勇. 电磁发射技术的研究现状与挑战[J]. 电工技术学报, 2023, 38(15):3943-3959.
	MA W M, LU J Y. Research status and challenges of electromagnetic emission technology[J]. Transactions of China Electrotechnical Society, 2023, 38(15):3943-3959. (in Chinese)
[4]	CORWEL C, ZOGHBI G, WEBB S, et al. Design and control of an underwater launch system[J]. IEEE Access, 2020,8:38633-38649.
[5]	郭灯华, 史铎林, 关晓存, 等. 电容驱动型多级感应线圈炮模型简化[J]. 电机与控制学报, 2022, 26(5):8-16.
	GUO D H, SHI D L, GUAN X C, et al. Simplified model of capacitor-driven multistage induction coilgun[J]. Journal of Electrical Machines and Control, 2022, 26(5):8-16. (in Chinese)
[6]	TAO X, WANG S H, HUANGFU Y P, et al. Geometry and power optimization of coilgun based on adaptive genetic algorithms[J]. IEEE Transactions on Plasma Science, 2015, 43(5):1208-1214.
[7]	NIU X B, AN Y Z, HU Y C. Exploration on Matching characteristics of slip and turns of multistage synchronous induction coilgun[J]. IEEE Transactions on Plasma Science, 2021, 49(2):928-933.
[8]	贾强, 关晓存, 龚想平, 等. 多级电磁感应线圈炮级间电磁力耦合分析[J]. 火炮发射与控制学报, 2023, 44 (1):88-93.
	JIA Q, GUAN X C, GONG X P, et al. Analysis of electromagnetic force coupling between gun stages of multi-stage electromagnetic induction coil[J]. Journal of Artillery Firing and Control, 2023, 44 (1):88-93. (in Chinese)
[9]	BAHARVAND M, KOLAGAR A D, PAHLAVANI M R A. Design,simulation,and parameter optimization of a multi-stage induction coilgun system[J]. IEEE Transactions on Plasma Science, 2021, 49 (7):2256-2264.
[10]	NIU X B, LI W Q, FENG J Y. Nonparametric modeling and parameter optimization of multistage synchronous induction coilgun[J]. IEEE Transactions on Plasma Science, 2019, 47(7):3246-3255.
[11]	张亚东, 肖刚, 王青子, 等. 基于正交试验法的同步感应线圈发射器关键设计参数选择[J]. 高电压技术, 2016, 42 (9):2843-2849.
	ZHANG Y D, XIAO G, WANG Q Z, et al. Selection of key design parameters of synchronous induction coil transmitter based on orthogonal test method[J]. High Voltage Technology, 2016, 42 (9):2843-2849. (in Chinese)
[12]	裘镓荣, 曾鹏飞, 邵伟平, 等. 基于PSO-LSSVM弹药装配质量预测方法[J]. 兵工学报, 2022, 43 (9):2379-2387.
	QIU J R, ZENG P F, SHAO W P, et al. A prediction method of ammunition assembly quality based on PSO-LSSVM[J]. Acta Armamentarii, 2022, 43 (9):2379-2387. (in Chinese)
[13]	陈子涵. 基于多模态Transformer的机电作动器剩余寿命预测[J]. 兵工学报, 2023, 44 (10):2920-2931. doi: 10.12382/bgxb.2022.0581
	CHEN Z H. Residual life prediction of electromechanical actuators based on multimodal transformer[J]. Acta Armamentarii, 2023, 44 (10):2920-2931. (in Chinese)
[14]	张通彤, 姜湖海, 岳巍, 等. 基于径向基函数神经网络的光电系统自适应控制[J]. 兵工学报, 2022, 43 (3):556-564. doi: 10.12382/bgxb.2021.0117
	ZHANG T T, JIANG H H, YUE W, et al. Adaptive control of optoelectronic system based on radial basis function neural network[J]. Acta Armamentarii, 2022, 43 (3):556-564. (in Chinese)
[15]	SHI D L, GUO D H. Research on control of instantaneous high power pulse energy supply of multi-stage coil transmitter based on deep learning[J]. Energy Reports, 2021(7):254-265.
[16]	田浩杰, 杨晓庆, 翟晓雨. 基于深度学习的线圈炮缺陷自动检测与分类[J]. 现代计算机, 2022, 28 (10):86-91.
	TIAN H J, YANG X Q, ZHAI X Y. Automatic detection and classification of coil gun defects based on deep learning[J]. Modern Computer, 2022, 28 (10):86-91. (in Chinese)
[17]	LE D V, GO B S, SONG M G, et al. Development of a capacitor bank-based pulsed power supply module for electromagnetic induction coilguns[J]. IEEE Transactions on Plasma Science, 2019, 47(5):2458-2463.
[18]	肖贞仁, 刘开培, 牛小波, 等. 电流环暂态模型在异步感应线圈发射器中的应用[J]. 电工技术学报, 2018, 33(17):3989-3997.
	XIAO Z R, LIU K P, NIU X B, et al. Application of current loop transient model in asynchronous induction coil transmitter[J]. Transactions of China Electrotechnical Society, 2018, 33(17):3989-3997. (in Chinese)
[19]	ZHOU L N, PAN S M, WANG J W, et al. Machine learning on big data:Opportunities and challenges[J]. Neurocomputing, 2017,237:350-361.
[20]	GUAN Z B, WANG J J, WANG X M, et al. A Comparative study of RNN-based methods for web malicious code detection[C]// Proceedings of the 2021 IEEE 6th International Conference on Computer and Communication Systems.Chengdu, China:IEEE,2021:769-773.
[21]	QIN T T, LIU M F, JI S Y, et al. Parameter weight analysis of synchronous induction electromagnetic coil launch system based on the entropy weight method[J]. IEEE Transactions on Plasma Science, 2024, 52(5):1865-1873.
[22]	万芯炜, 王晶, 杨辉, 等. BP神经网络结合粒子群优化卡尔曼滤波的MEMS陀螺随机误差补偿方法[J]. 兵工学报, 2023, 44 (2):556-565. doi: 10.12382/bgxb.2022.0110
	WAN X W, WANG J, YANG H, et al. BP neural network combined with particle swarm optimized Kalman filtering for random error compensation of MEMS gyro[J]. Acta Armamentarii, 2023, 44(2):556-565. (in Chinese)
[23]	刘芳, 李士伟, 卢熹, 等. 基于PSO-CNN-XGBoost水下柱形装药峰值超压预测[J]. 兵工学报, 2024, 45(5):1602-1612. doi: 10.12382/bgxb.2023.0743
	LIU F, LI S W, LU X, et al. Prediction of peak overpressure of underwater column charge based on PSO-CNN-XGBoost[J]. Acta Armamentarii, 2024, 45(5):1602-1612. (in Chinese)
[24]	JIAO K R, SHEN X J.Real-time evaluation model construction of GIS conductor thermal state based on PSO-BP neural network[C]//Proceedings of the 2023 3rd International Conference on Electrical Engineering and Control Science (IC2ECS). Hangzhou,China:IEEE,2023:1690-1695.

参数	数值
线圈长度/mm	102
线圈内径/mm	63
线圈外径/mm	177
线圈匝数	55/55/50
电容器放电电压/kV	5/7
电容值/μF	1000
驱动线圈材料	铜
身管长度/mm	1200
身管直径/mm	50

参数	数值
线圈长度/mm	102
线圈内径/mm	63
线圈外径/mm	177
线圈匝数	55/55/50
电容器放电电压/kV	5/7
电容值/μF	1000
驱动线圈材料	铜
身管长度/mm	1200
身管直径/mm	50

参数	数值
内半径/mm	10
径向厚度/mm	4.5
轴向长度/mm	135
质量/g	447
初速度/(m·s^-1)	0
材料	铝

参数	数值
内半径/mm	10
径向厚度/mm	4.5
轴向长度/mm	135
质量/g	447
初速度/(m·s^-1)	0
材料	铝

电压/kV	仿真结果/ (m·s^-1)	实验结果/ (m·s^-1)	误差/ (m·s^-1)
5	40.0	40.4	0.4
7	78.1	78.6	0.5