Diesel engine cylinder pressure recovery based on vibration signals
ZHANG Shuai1,2,ZENG Ruili3
1.Military Representative Office in Xi'an，Military Representative Bureau of Munitions，Xi'an 710000，China；
2.Postgraduate Training Brigade, Military Transportation University, Tianjin 300161, China;
3.Military Vehicle Department，Military Transportation University, Tianjin 300161, China
Abstract：As an important index of a diesel engine, its cylinder pressure directly reflects its combustion state.Due to complex operating conditions of the engine, it is in a non-stationary state under most conditions.How to improve the cylinder pressure’s recognition accuracy under multiple working conditions becomes the key of cylinder pressure recovery.Here, a new method for cylinder pressure recovery using vibration signals was proposed.Taking the maximum entropy spectral density of a vibration signal as its feature, Douglas-Puke algorithm was adopted to reduce the dimension number of input and output feature vectors.Finally, the BP neural network with multi-hidden layer optimized using the genetic algorithm was used to effectively restore the cylinder pressure curve of the engine under multiple operating conditions.The test results showed that after the average treatment, the maximum restoring error of the cylinder pressure curve’s peak values is 0.05 MPa and their maximum location error is 0.6 °CA to meet the accuracy requirement of cylinder pressure recovery.
[1] R. Johnsson. Indirect measurements for control and diagnostics of IC engines[D]. Ph.D. Thesis, Luleå University of Technology,Sweden, 2004.
[2] A. Cocchi, P. Andreini, L. Cassitto, et al. Accelerometer Based Methodology for Combustion Parameters Estimation[J]. Energy Procedia, 2015, 7(81):950-959.
[3] K. Bizon, G. Continillo, E. Mancaruso, et al. Reconstruction of In-Cylinder Pressure in a Diesel Engine from Vibration Signal Using a RBF Neural Network Model[J]. SAE Technical Paper 2011-24-0161, 2011.
[4] R. Johnsson, A. Ågren. Prediction of points and tendencies of the pressure waveform from crankshaft speed measurements[J]. Kerntechnik, 2009, 74(1):51-54.
[5] F. Taglialatela, M. Lavorgna, E. Mancaruso, et al. Determination of combustion parameters using engine crankshaft speed[J]. Mechanical Systems & Signal Processing, 2013, 38(38):628–633.
[6] C. Bennett, J.F. Dunne, S. Trimby, D. Richardson. Engine cylinder pressure reconstruction using crank kinematics and recurrently-trained neural networks[J]. Mechanical Systems and Signal Processing, 2017,85:126-145.
[7] Y. Ren. Detection of knocking combustion in diesel engines by inverse filtering of structural vibration signals[D]. Ph.D. Thesis, University of New South Wales, Australia, 1999.
[8] 姚建军,向阳,周勇,等.基于倒频谱方法的柴油机气缸压力识别[J].船海工程,2006,35(1):37-39.
J. J. Yao, Y. Xiang, Y. Zhou, et al. Identification of cylinder pressure of diesel engine based on the cepstral analysis[J]. Ship & ocean engineering, 2006, 35(1):37-39.
[9] Y. Gao, R.B. Randall. Reconstruction of diesel engine cylinder pressure using a time domain smoothing technique[J]. Mechanical Systems and Signal Processing, 1999, 13(5): 709–722.
[10] V. G. Zurita. Vibration based diagnostics for analysis of combustion properties and noise emissions of IC engines[D]. Ph.D. Thesis, Luleå University of Technology, Sweden, 2001.
[11] H. Du, L. Zhang, X. Shi. Reconstructing cylinder pressure from vibration signals based on radial basis function networks[J]. Proceedings of the Institution of Mechanical Engineers Part D Journal of Automobile Engineering, 2001, 215(6):761-767.
[12] B. Colin. Reconstruction of gasoline engine in-cylinder pressures using recurrent neural networks[D]. University of Sussex, 2014.
[13] 高洪滨, 欧阳光耀, 张萍. 基于BP神经网络的柴油机气缸压力重构[J]. 内燃机工程, 2005, 26(1):68-70.
H.B. Gao, G.Y. OuYang, P. Zhang. Reconstruction of Diesel Engine Cylinder Pressure Based on BP Neural Network[J]. Chinese Internal Combustion Engine Engineering, 2005, 26(1):68-70.
[14] R. Johnsson. Cylinder pressure reconstruction based on complex radial basis function networks from vibration and speed signals[J]. Mechanical Systems & Signal Processing, 2006, 20(8):1923-1940.
[15] K. Bizon, G. Continillo, E. Mancaruso, et al. Towards On-Line Prediction of the In-Cylinder Pressure in Diesel Engines from Engine Vibration Using Artificial Neural Networks[J]. SAE Technical Paper 2013-24-0137, 2013.
[16] K. Bizon, G. Continillo, S. Lombardi, et al. ANN-based Virtual Sensor for On-line Prediction of In-cylinder Pressure in a Diesel Engine[C]. Proceedings of the 24th European Symposium on Computer Aided Process Engineering, Hungary, 2014, 33:763-768.
[17] S. Trimby, J. F. Dunne, C. Bennett, et al. Unified approach to engine cylinder pressure reconstruction using time-delay neural networks with crank kinematics or block vibration measurements[J]. International Journal of Engine Research, 2016, 18(3):1-39.
[18] 常春, 贾继德, 曾锐利,等. 基于时-频相干与 RBF 网络的气缸压力识别研究[J]. 车用发动机, 2016, 39(5):87-92.
C. Chang, J. D. Jia, R. L. Zeng, et al. Recognition of Cylinder Pressure Based on Time-frequency Coherence and RBF Network[J]. Vehicle engine, 2016, 39(5):87-92.
[19] J. P. Burg. The Relationship Between Maximum Entropy Spectra and Maximum Likelihood Spectra[J]. Geophysics, 1972, 37(2):375-376.
