BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//6.6.4.4//EN TZID:Asia/Jerusalem X-WR-TIMEZONE:Asia/Jerusalem BEGIN:VEVENT UID:0-399@aerospace.technion.ac.il DTSTART;TZID=Asia/Jerusalem:20180226T163000 DTEND;TZID=Asia/Jerusalem:20180226T173000 DTSTAMP:20230530T180352Z URL:https://aerospace.technion.ac.il/events/maximum-conditional-probabilit y-stochastic-controller-for-linear-systems-with-additive-cauchy-noises/ SUMMARY:Maximum Conditional Probability Stochastic Controller for Linear Sy stems with Additive Cauchy Noises DESCRIPTION:Lecturer:Nati Twito\n Faculty:Department of Aerospace Engineeri ng\n Institute:Technion – Israel Institute of Technology\n Location:Clas sroom 165\, ground floor\, Library\, Aerospace Eng.\n Zoom: \n Abstract: \ n Details: \n The majority of practical estimation and control solutions a re based on system models with additive Gaussian noises. Gaussian distribu tion\, being light tailed\, does not capture significant fluctuations that occur in many engineering applications\, such as atmospheric noises and a ir turbulence\, underwater acoustic noises and image processing. It was sh own that those noises are better described by heavy tailed distributions t hat exhibit significant impulsive characteristics. The heavy tailed distri butions are characterized by probability density functions with tails that are not exponentially bounded. Named after the French mathematician\, Aug ustin Louis Cauchy\, the heavy-tailed Cauchy distribution has been shown t o much better represent this type of fluctuations. The challenge of using this distribution is that it does not have a moment generating function. S pecifically\, its first moment is not well defined and its second and high er moments are infinite.\nIn the recent years\, progress was reported in t he area of estimation of linear systems with additive Cauchy noises. As pa rt of its solution\, the estimator computes explicitly the conditional pro bability density function (pdf) of the state given the measurement history \, or its characteristic function. Those were used to derive a stochastic optimal-predictive controller for the system. Although demonstrating good performance characteristics\, the proposed controller was shown to entail high numerical complexity. This motivated the alternative approach present ed in the current study.\nIn this work a stochastic controller\, inspired by the sliding mode control methodology\, is proposed for linear systems w ith additive Cauchy distributed noises. The design goal is to maximize the prior probability of the system state or its linear combination to be wit hin a given bound around the regulation point. The control law utilizes th e time propagated pdf of the system state given measurements that is compu ted by the Cauchy estimator. The single-state controller was derived using two equivalent implementations: one that relies directly on the above men tioned prior pdf while the second uses the characteristic function of that pdf. The latter was addressed mainly because in the multi-state case only the characteristic function can be determined by the respective Cauchy es timator. The controller performance was evaluated numerically\, and compar ed to an alternative approach presented recently and to a Gaussian approxi mation to the problem. A fundamental difference between the Cauchy and the Gaussian controllers is their response to noise outliers. While all contr ollers respond to process noises\, even to the outliers\, the Cauchy contr ollers drive the state faster towards zero after those events. On the othe r hand\, the Cauchy controllers do not respond to measurement noise outlie rs\, while the Gaussian does. The newly proposed Cauchy controller exhibit s similar performance to the previously proposed one\, while requiring low er computational effort\, and is much easier to implement. CATEGORIES:Seminars LOCATION:Classroom 165\, ground floor\, Library\, Aerospace Eng. END:VEVENT BEGIN:VTIMEZONE TZID:Asia/Jerusalem X-LIC-LOCATION:Asia/Jerusalem BEGIN:STANDARD DTSTART:20171029T010000 TZOFFSETFROM:+0300 TZOFFSETTO:+0200 TZNAME:IST END:STANDARD END:VTIMEZONE END:VCALENDAR