The Research On Coherent Acoustic Field Noise Prediction Theory And Its Application In HVDC Transmission Systems

High Voltage Direct Current (HVDC) transmission has been developing rapidly in network system with its advantages of long distance transmission, large capacity, and low power loss. With the improving of national environmental awareness in recent years, the impact of radiation noise has been the significant issue that must be taken into consideration in Ultra High Voltage (UHV) DC transmission engineering because of the large radiation noise of DC converter station. Due to the high cost of later noise control for established engineering, it is very important to evaluate the noise level by calculation in project designing and make prediction for noise radiation. At present, several foreign noise prediction software have been used to predict the noise radiation of the HVDC project and have met the expectation of noise control evaluation, both in the early and late stage. However, most of the currently used outdoor noise prediction software are based on ISO9613-2standard, which ignores the coherent effects of the sources. The dominant sound strength at converter station mainly concentrates in the low and mid frequency range, and many units of the same model are used in DC station, such as transformer, capacitor, etc. There is strong coherent acoustic field in the surrounding of converter station, making the outdoor noise prediction algorithms based on energy method insufficient because of large deviation in noise prediction for certain areas.To deal with the disadvantage of current algorithms, this paper has studied the factors which influenced the noise prediction accuracy of HVDC converter station deeply, and proposes an engineering outdoor noise prediction algorithm model based on the theory of coherent acoustic field in combination with the geometric diffraction algorithm and coherent image source method. The coherent prediction model can not only calculate coherent acoustic field by multiple paths diffraction of sound when the sound source go through a finite length barrier or multiple barrier, but also calculate coherent acoustic field by multiple reflections after the sound source reflected repeatedly between interface such as barriers. The computational efficiency of the coherent prediction model was improved greatly on account of the engineering simplified formula of diffraction acoustic field, and avoided the complex integral calculation of diffraction sound theoretical value. By applying the model to calculate sound field of a HVDC transmission project, the calculated value is compared between ISO9613,BEM and the measured values. The results show that the proposed model can predict the surrounding coherent acoustic field and the calculated value is more closer to the test result than ISO9613model.The main work of this paper lies in the following aspects:(1) It reviews the current study method and result of noise prediction models for HVDC systems, analyzes the factors related to the accuracy of prediction and analyzed the noise radiation characteristics of main devices used in DC stations. The main sound sources and dominant frequencies of d coherent acoustic field are clarified.(2) It expounds currently used the Fresnel method and geometric diffraction (GTD) in barrier attenuation algorithm, simplifies the Hadden-Pierce theoretical solution for single rigid thin barrier diffraction in the GTD, and proposes a simplified algorithm of diffraction acoustic field which is suitable for engineering calculation. This algorithm simplifies the calculation process and reduces the computing time greatly by avoiding the complex numerical calculation in the original method and singular integral process. The calculation method can also determine the diffraction at any position or direction near the edge of diffraction field, including acoustic shadow boundary barrier and area very close to diffraction side.(3) It proposes an outdoor half-open space coherent source model to calculate the inside and outside sound field excited by the point source in the half-open space formed by the ground and two or three vertical barrier of finite length. Compared against existing prediction method and boundary element method, the proposed imaginary source model has higher precision in wide frequency range. It also considers the phase of the reflected wave in terms of interface impedance and the interference effect between different sound waves in PLD. Compared with the plane wave reflection coefficient given in Lemire coherent image source method, the current coherent image source model determines the actual path with ray-tracing technique. The reflection coefficient is calculated from positive incidence angle, which is a more consistent to reality. This model is especially suitable for predicting the external half-open sound field, and also can be used for prediction with sound absorbing interface.(4) It designs a scaled model of a converter-transformer box wall. Modal tests are made and amplitude and phase under sine excitation of different frequencies are obtained for multiple points on the surface. The experimental results show that the vibration shapes are complex due to the structural characteristics. For excitation frequency which reaches seventh order frequence, the randomness of phase and amplitude of vibration is obvious. In conclusion, a single transformer in one direction can be divided into a number of equivalent point sources, and calculation can be performed assuming irrelevant sources. For the transformers of different phases in a group of converter transformers (consisting of three or six transformers), the equivalent sound sources at same position of each transformer should be modeled as sources different from each other by120°and coherent sound sources must be used.(5) Based on the outdoor Lemire half-open space coherent source model and combined with the conclusion of the scaled converter transformer model test, an acoustical model for noise prediction of converter transformer group. This model takes the influence of sound source and the phase during propagation into account. From comparison against the boundary element method model and actual test results, the model can accurately predict in converter transformer coherent acoustic field with higher precision than current standard ISO9613-2models. With coherent acoustic field interference effect taken into consideration, it improves the outdoor noise prediction accuracy, thus making it an important contribution to the noise prediction for HVDC station.(6) This paper summarizes the main research results and points out that the problem which needs to be further studied and solved.