A principal consideration on general Bayes'filters for a state estimation of environmental soun

• 作 者: M.OHTA E.UCHINO
• 机 构: Department of Electronic and Information Engineering,Kinki University,Takaya,Higashi-Hiroshima 729-1
• 出 处: Chinese Journal of Acoustics
• 发表日期: 1992年02期
• 关键词: connected quantization giving noisy stationary viewpoint successive truncation a

This paper describes a fundamental consideration on our workson the design of general Bayes’filters for the state estimation ofnon-stationary,non-linear,and non-Gaussian environmental sound andvibration systems.We have discussed an essential point of several Bayes’filters proposed by using the orthogonal or non-orthogonal expansionform of Bayes’theorem.They can estimate any kinds of statistics ofarbitrary function type of state variables including the lower and thehigher order statistics connected with the L_x evaluation index in theenvironmental sound and vibration systems.Here,we have mainlyfocussed on giving the fundamental viewpoints of their design policies.Some new estimation methods and new results not yet published areincluded.

This paper describes a fundamental consideration on our works of the design of general Bayes ’filters for the state estimation of non-stationary, non-linear, and non-Gaussian environmental sound and vibration systems. We have discussed an essential point of several Bayes’ filtered by using the orthogonal or non-orthogonal expansionform of Bayes’ theorem. They can estimate any kinds of statistics ofarbitrary function type of state variables including the lower and thehigher order statistics connected with the L_x evaluation index in the environmental sound and vibration systems. Here, we have mainlyfocussed on giving the fundamental viewpoints of their design policies.Some new estimation methods and new results not published areincluded.