Research And Application Of Near - Field Noise Calculation Model For Substation

According to the analysis of noise data in substations, the noise around the transformers and other noise sources can reach more than80dB. It is well known that chronic exposure to high intensity noise may cause hearing damage. In order to protect the employee’s health, avoid the hearing impairment due to contact with noise sources, the power company need to predict near-field noise output when the design of substation projects. Current mainstream noise prediction software is to focus on far-field noise calculation model and the near-field noise calculation model was relatively weak. Therefore, this study designs a computation model that for substation near-field noise attenuation calculating and applies it to software.Based on the analysis of some forecasting software at home and abroad, and the comparison of calculation method for noise attenuation, this paper presents a new computation module that applies to substation near-field noise attenuation. In this model takes into account a variety of factors that not only geometric divergence, also joined the atmospheric absorption, diffraction when across the barrier, surface absorption and other factors in calculating noise attenuation process in order to reduce deviation between model and the real world.When calculating the geometric propagation of noise attenuation,we find common models are not suitable for application in substation near-field noise by analyzing common computational model. Therefore, based on the theory of acoustic wave propagation, this paper builds the new geometric model of near-field noise attenuation calculation about the main noise sources in the substation. Through numerous experiments, verifies the correctness of the model and achieves the application sphere of the model.Based on establishment of mathematical model, use the.NET technology and object-oriented programming, we develop the software of substation near-field noise prediction system. And elaborate the information data of the noise model, the structure of the software module, the algorithms, data structures, and other aspects. In addition, compares the software predicted data with experimental data, the calculated data of Cadna/A Software and find that the software is modeling faster, calculate more accurate, user friendly features.