| NVH (Noise Vibration and Harshness) is one of the most important indexes to measure the vehicleperformance, it directly affects the vehicle’s comfort quality, and the vehicle’s comfort directly influencethe market competitiveness. Study and Control of noise and vibration characteristics of the vehicle lies inthe control of internal combustion engine noise, it need to carry out the source identification of internalcombustion engine noise. In order to control the noise of internal combustion engine, it should studyvibration and noise of internal combustion engine, which is used to identify the source of vibration andaccurately measure and analyze vibration sources, such as the source type, the frequency, the soundpressure level and so on. According to the analysis results of the noise and vibration source, it takescorresponding measures and reduces the engine sound pressure level. For nearly three decades, technologyworkers unremittingly are devoted to the study design, low noise combustion system, engine noiseisolation and damping and noise reduction technology of noise optimization, and already made manybreakthroughs. Relevant data shows that the overall noise level was decreased by12dB compared with theoverall level of noise radiation of thirty years ago.Even so, the engine noise radiation levels still have falling space and necessity. Since the engine is acomplex mechanical system, when the engine system is running, its internal multiple components aresimultaneously operating. This leads to response of the target component is affected by non-objectionparts vibration signal, the spectrum overlap, experimental means to measure the response is very complex,and has the high cost,which makes the traditional identification of noise source has the limitation. With theapplication of new theories and new technologies, the use of modern signal processing methods to studythe characteristics of the engine noise and vibration has important theoretical and practical value.In view of the above problems, this paper carried out the corresponding research work:Firstly, this paper explains the research background, harmfulness of the diesel vibration and noisecharacteristics, as well as the current processing technical problems in the vibration signals of diesel engine.And expounds the development process and research status of blind separation algorithm, and the vibrationsignal of internal combustion engine modern treatment methods, and points out that the time domain blindidentification algorithm’s value in the field of internal combustion engine noise source identification.Secondly, by using the continuous wavelet transform on the engine surface vibration signaltime-frequency analysis, and by decomposing the characteristic of each parts of engine surface vibration acceleration signal to identify the main noise source of engine in the process of operation, it provides somemethods and approaches for the identification of internal combustion engine noise source;Thirdly, this paper establishs a linear source signal instantaneous mixing model and the source signallinear convolutive mixture model, presents the basic concepts of information theory, introduces thecommonly used algorithm solution separation of linear instantaneous mixed blind, and describes in detailthe process of FastICA algorithm, the numerical results show the efficiency and the robustness of theFastICA algorithm. At the same time, this paper also introduces the commonly used mixed algorithm forsolving linear convolutive blind source separation problem, and detailed description of the MBLMSalgorithm, and verifies the efficiency and the robustness by the numerical results.Finally, the FastICA algorithm and the improved MBLMS algorithm was used for mixed signalseparation from diesel engine surface. The improved MBLMS algorithm is verified high efficiency andreliability by comparing separation results, the further research of non-stationary vibration of internalcombustion engine noise is studied. |
PDF Full Text Request