Research On Active Noise Control For Engine Exhaust Pipe Of Legged Robots

Acoustic noise problems become more and more evident as increased numbers of industrial equipment such as engines, blowers, fans, transformers, and compressors in use. At present, ANC(active noise control) technology has become research highlights at home and abroad, which has properties of smaller volume, higher efficiency and lower cost, etc. The technology has advantages over passive noise control and it improves and makes up for the deficiency of the passive form. It develops very fast especially for robot system powered by engine. Pipeline structure and placement of acoustic components has a significant effect on the noise reduction for engine of quadruped robot. How to make use of the simulation analysis of pipeline structure and find a proper installation site is a key point in this topic. With rapid development of ANC algorithm, how to design an experimental platform for ANC system and choose a suitable algorithm of adaptive control is a difficult point for study. This article has carried on the structure simulation and acoustic field simulation to guide the design of structure and layout of acoustic components. Through MATLAB simulation, results of the off-line identification and online identification for ANC system is compared. An experiment platform used to verify the effectiveness of the adaptive control is established and has obtained better noise reduction for engine.Firstly, simulation for acoustic field and pipeline structure is analyzed faced to onedimensional pipe. Gas flow analysis and steady-state heat and solid coupling analysis are achieved. Bias layout model is little affected by temperature and flow velocity and is more fit for placement of acoustic components. And then through analysis for acoustic simulation, pneumatic noise and vibration coupling for pipe is calculated and the results prove that it is unnecessary to consider the effects of acoustic vibration coupling.Secondly, in order to simulate the pipe noise of engine, spectral analysis is carried out for noise of engine. There are two basic structures for ANC system, the feed-forward and feed-back control system. And through MATLAB, emphatically on-line and off-line system identification based on FXLMS algorithm are compared to decide which is better. Feedforward control strategy for ANC based on off-line identification is selected to reduce the noise of engine. Factors that affect the performance of stability and convergence for FXLMS algorithm are discussed, which can provide the basis for the regulation of experimental parameters.At last, in order to test the actual effect of noise reduction by FXLMS algorithm, an experimental platform is carried out by using LABVIEW as the upper machine to implement ANC system. Accurately identifying the secondary path lays a foundation to apply to ANC system. By Face to Face platform, the feasibility of ANC system is confirmed. On bias layout platform, ANC system based on FXLMS algorithm is verified that it can realize the noise reduction for engine.