FEASIBILITY STUDY OF ACTIVE CONTROL OF NOISE PERCEIVED BY OPERATORS OF LARGE AGRICULTURAL MACHINES

A simulator for active noise control (ANC) in an enclosed space is presented in this paper. The simulator is used to predict as accurately as up to date numerical techniques allow, the effect of ANC in the driver's cabin of large agricultural machines. The acoustical problem is approximated using the finite difference time domain technique, which allows simultaneous analyses of the digital electronics used for ANC. To our knowledge this combined approach is unique in the field of theoretical studies of ANC. As an example of the use of the simulator, this paper focuses in particular on a feed-forward control mechanism using a parallel set of filters, optimized by an integrating curve fit algorithm. Results include the influence of algorithm parameters and uncorrelated noise on convergence and highlight the physical mechanisms involved in the control of the sound field.