A detailed discussion of the method and its evolution in the past decade as well as an efficient implementation of it … )y�A9D�=Bb�3nl��-n5�jc�9����*�M��'v��R����9�QLДiC�r��"�E^��;.���`���D^�a�=@c���"��4��HIm���V���%�fu1�n�LS���P�X@�}�*7�: We then analyze Kalman filtering techniques for nonlinear systems, specifically the well-known Ensemble Kalman Filter (EnKF) and the recently proposed Polynomial Chaos Expansion Kalman Filter (PCE-KF), in this Bayesian framework and show how they relate to the solution of Bayesian inverse problems. The text incorporates problems and solutions, figures and photographs, and astonishingly simple derivations for various filters. The standard Kalman lter deriv ation is giv :f��'� p���9�H��MMp����j����:���!�7+Sr�Ih�|���I��ȋ<
}+��q�������ǜҟ~�H�����u�\���3���0N���f�A���5W��Oy�z_�@�ZJb|V��� �B4�\Jˣ�5~G7���/O�{�6�+�J�5�a��R�/���� �,um��f������l�ZfW�B�)0��u5w"I���s�b���{�R0�M�0�Y{W,Τ�}���[���,�m��@�B羾s"� iՍ��n��{)�nHC��v�˦�濹�V�ÄڜU7�����H8
��BpK���)h����S,嗟�U�j�j0_�< ������2�Y��H&�(��s The solution sec-tion describes the two key computational solutions to the SLAM problem through the use of the extended Kalman filter (EKF-SLAM) and through the use of Rao-Blackwellized par-ticle filters (FastSLAM). 8.4.2 Kalman-Schmidt Consider Filter / 325 8.5 Steady-State Solution / 328 8.6 Wiener Filter / 332 8.6.1 Wiener-Hopf Equation / 333 8.6.2 Solution for the Optimal Weighting Function / 335 8.6.3 Filter Input Covariances / 336 8.6.4 Equivalence of Weiner and Steady-State Kalman-Bucy Filters / … <> Filtering noisy signals is essential since many sensors have an output that is to noisy too be used directly, and Kalman filtering lets you account for the uncertainty in the signal/state. I know that amcl already implements particle filter and you can use kalman filter with this package, but the problem with them is that amcl needs robot's initial position. A Kalman filter is an optimal estimator - ie infers parameters of interest from indirect, inaccurate and uncertain observations. ; difficulty (3) disappears. Looking on internet I saw the two solutions are particle and kalman filter. ��FIZ�#P��N����B
o�9Ж]�K�4/.8�X��x:P�X��q��
��?Y���'��2yQmw��L\�N�9--^�BF? �{hdm>��u��&�� �@���ŧ�d���L\F=���-�ӫ>��X��ZF[r��H��2f���$�7x���Kˉl� �"�j��\p� �cYz4I�+-�Y��Ȱ����IL�í
����]A��f�|ץ��{��o:CS83�����鋳$��e��%r�b��`� ��� �L���c$�p^�����>yKXˑ�!�QX��1S�y�+ N�k�
TP��FKV@�xZ��Q�KF씈lh�M�h��{6�E�N����Kz^���ؕ���)�@Z̮'�}�Fd�7X)�U2Yu�G��
6IQI9s���@�����W�TtK�=�r�:�S)e�3Q1ʫcGc�qxIP�|� }āpgm���N'\�&��j��؊oE�`G|����d�yd?�q,H|P����2y�':r�X�k��xI�@��^��?�ʪ�]�
��μ��2�C@ol�!�/. �����C A time-invariant Kalman filter performs slightly worse for this problem, but is easier to design and has a lower computational cost. "r\�����S�j��_R('T0��! Unlike static PDF Kalman Filtering: Theory and Practice Using MATLAB 3rd Edition solution manuals or printed answer keys, our experts show you how to solve each problem step-by-step. 2 History matching with the ensemble Kalman filter The EnKF was first introduced by Evensen [11] in 1994 as a way to extend the classical Kalman filter to nonlinear problems [12]. The quaternion kinematic equation is adopted as the state model while the quaternion of the attitude determination from a strapdown sensor is treated as the measurement. However, in practice, some problems have to be solved before confidently using the Kalman filter. A , B And C Are The Matrices To Use For The State Space Blocks . Kalman Filter Extensions • Validation gates - rejecting outlier measurements • Serialisation of independent measurement processing • Numerical rounding issues - avoiding asymmetric covariance matrices • Non-linear Problems - linearising for the Kalman filter. Its use in the analysis of visual motion has b een do cumen ted frequen tly. ��/;��00oO��� ��Y��z����3n�=c�ήX����Ow�;�߉v�=��#�tv��j�x�S b ~����h���L��hP�Qz1�ߟѬ�>��
$��ck3Y�C��J �z�=����� 1 The Discrete Kalman Filter In 1960, R.E. The teaching assistants will answer questions in office hours and some of the problems … It is used in a wide range of engineering and econometric applications from radar and computer vision to estimation of structural macroeconomic models, and is an important topic in control theory and control systems engineering. W��zܞ�"Я��^�N�Q�K|&�l �k�T����*`��� %�쏢 Together with the linear-quadratic regulator (LQR), the Kalman filter solves the linear–quadratic–Gaussian controlproblem (LQG). I am looking for the solution for problem 2 kalman filter equation to implement it. �-���aY��k�S�������� ii ABSTRACT TREND WITHOUT HICCUPS - A KALMAN FILTER APPROACH By ERIC BENHAMOU, PhD, CFTe, CAIA, CMT DATE: April 2016 Have you ever felt miserable because of a sudden whipsaw in the price that triggered an. The model … Kalman published his famous paper describing a recursive solution to the discrete-data linear filtering problem. 11.1 In tro duction The Kalman lter [1] has long b een regarded as the optimal solution to man y trac king and data prediction tasks, [2]. The Kalman filter is an efficient recursive filter that estimates the internal state of a linear dynamic system from a series of noisy measurements. In 1960, R.E. Kalman Filter T on y Lacey. It is common to have position sensors (encoders) on different joints; however, simply differentiating the posi… problems for linear systems, which is the usual context for presenting Kalman filters. ��$$���ye��:�&�u#��ς�J��Y�#6
��&��/E@\�[b6c��!�w�LH�����E'���ݝ}OVe�7��"��wOh{�zi by�k���Hʗ;��d�E���Hp,�*�ڵb�pX�X�On%*�w+lS�D��t����E7o۔�OOܦ������fD������.� n��L�2":��Z��zo���x0��S�1 xI��J!K##���L���As�G�@�� "��`6��X9A`�*f����ޫ9LTv!�d(�2!=
���v�Mq����*��n��X{��.g@���W�wZ=�2 Ό> ��|�s�*��}�=:B�@���L�w�!8w;s���^�mdʿ��", �PST�j\}*�[�, �7����U��U���L�jCw���g����k�c�O���,*���#����c��p��~�R*�����V�@�����}�M� �\�a��b}��2��l�1G�Ai�~Z��9�.�fQ>�nZC@"�`$�C;�����������B�f���E5����p���{O��kk%��*R�7���Dͧu"����=��ڳ2����d�}�i\9�Ʈ����F�[E�C��`�������5[���ޢ���>:�'��9��9�L;���f�=�,*�Ā��8����^U�/Z2�{l6|wu�;� D D�z�#��y>>|\w���Є:O�c�7i��6T���K�~2#p�+�`�ov�.x�Fڷ��´��/+���/�T���v���y��x�FZ�G`9hri����A[�{f�Ə�?,������]�_�3�_��f��5p ���7�=W 2. �� �Л���1lNK?����D���J�)�w� *-���Òb�^i`#yk.�a>\�)���P (l�
V���4���>���Fs3%���[��*ӄ[����K=Dc�h����2�^�'^���zԑD3R�� *�
�)��u��Z�ne�����}���qg����}��Ea(�� %PDF-1.2 uǩ���F��$]���D����p�^lT�`Q��q�B��"u�!�����Fza��䜥�����~J����Ѯ�L���
��P�x���I�����N����� �Sl.���p�����2]er 9S��s�7�O These problems are related both with the numerical accuracy of the algorithm proposed by Kalman, and with the estimation of parameters that in the conventional Kalman filter are assumed to be known. With the state-transition method, a single derivation covers a large variety of problems: growing and infinite memory filters, stationary and nonstationary statistics, etc. Gauss (1777-1855) first used the Kalman filter, for the least-squares approach in planetary orbit problems. This question hasn't been answered yet Ask an expert. d��zF��y��`���ȏV�Ӕ_�'����SQ4����t����=�_]��ڏ�|�͞�f$�O|��u������^�����-���Ն���QCy�c^�ؘ�9��}ѱit��ze���$�=��l �����j��
�.�k�±'�2�����n��ͅg��I����WE��v�����`mb�jx'�f���L|��^ʕ�UL�)��K!�iO��薷Q/��ݲ�:E�;�A�رM�.� ���� �I��¯;��m:�(�v� ���^k�5`�_Y��8 �B�[Y!�X�-2[Ns��. Certain approximations and special cases are well understood: for example, the linear filters are optimal for Gaussian random variables, and are known as the Wiener filter and the Kalman-Bucy filter. � Kalman published his famous paper describing a recursive solution to the discrete-data linear filtering problem [Kalman60]. e��DG�m`��?�7�ㆺ"�h��,���^8��q�#�;�������}}��~��Sº��1[e"Q���c�ds����ɑQ%I����bd��Fk�qA�^�|T��������[d�?b8CP� Kalman ﬁlters divergence and proposed solutions Laura Perea - Institut de Ci`encies de l’Espai (CSIC-IEEC) November 22, 2006 Abstract This research was motivated by the problem of determining relative orbit positions of a formation of spacecrafts. x��\Ks�v��������h'x?�JU��q�R��T*�u�(Y�-�z�r�_��0h�`f�4m�\*�3��ϯ
�܈An������~��ͽ�oO^������6����7�JZ�9��D��қ!��3b0������ǻ��7�l���� �����P;���o|ܾ���`��n�+a��w8��P;3� ��v�Zc�g;
�:g����R��sxh�q2��o/��`/��O��*kM�
��Y��� Since that time, due in large part to advances in digital computing, the Kalman filter has been the subject of extensive research and application, particularly in the area of autonomous or assisted navigation. In 1960, Kalman published his famous paper describing a recursive solution to the discrete-data linear filtering problem. We will make sets of problems and solutions available online for the topics covered in the lecture. ��W���PF(g@���@.���E�oC)�e(3ֳ��0�N It is the student's responsibility to solve the problems and understand their solutions. Welch & Bishop, An Introduction to the Kalman Filter 2 UNC-Chapel Hill, TR 95-041, July 24, 2006 1 T he Discrete Kalman Filter In 1960, R.E. 17 0 obj The solution, however, is infinite-dimensional in the general case. The Kalman filter is the natural extension of the Wiener filter to non-stationary stochastic systems. In estimation theory, Kalman introduced stochastic notions that applied to non-stationary time-varying systems, via a recursive solution. $�z�oظ�~����L����t������R7�������~oS��Ճ�]:ʲ��?�ǭ�1��q,g��bc�(&���
e��s�n���k�2�^g
�Q8[�9R�=;ZOҰH���O�B$%��"�BJ��IF����I���4��y���(�\���^��$Y���L���i!Ƿf'ѿ��cb���(�D��}t��ת��M��0�l�>k�6?�ԃ�x�!�o\���_2*�8�`8������J���R⬪. C.F. In such a problem, ... Kalman Filter … Kalman filtering is used for many applications including filtering noisy signals, generating non-observable states, and predicting future states. Notes on Kalman Filtering Brian Borchers and Rick Aster November 7, 2011 Introduction Data Assimilation is the problem of merging model predictions with actual mea-surements of a system to produce an optimal estimate of the current state of the system and/or predictions of the future state of the system. I've seen lots of papers that use Kalman Filter for a variety of problems, such as noise filtering, sub-space signal analysis, feature extraction and so on. The block uses a time-varying Kalman filter due to this setting. Having guessed the “state” of the estimation (i.e., filtering or prediction) problem In real-life situations, when the problems are nonlinear or the noise that distorts the signals is non-Gaussian, the Kalman filters provide a solution that may be far from optimal. It's easier to figure out tough problems faster using CrazyForStudy. The Q matrix is time-varying and is supplied through the block inport Q. %PDF-1.4
%����
SLAM is technique behind robot mapping or robotic cartography. H��Wɒ����WԱ� 1��ɶ,K>)B1�i��"Y� �=�߰��]�̪�e��h ��\^�|�����"�ۧZD��EV�L�χ�ь�,c�=}��ϱ؍OQE1�lp�T�~{�,;5�Պ�K���P��Q�>���t��Q
��t�6zS/&�E�9�nR��+�E��^����>Eb���4����QB'��2��ѣ9[�5��Lߍ�;��'���: s��'�\���������'{�E�/����e6Eq��x%���m�qY$���}{�3����6�(݇�
�~m= ˗��JO��bN�7��C�5��$��S�P��hà��zl�f����ns���I���1,�ͅ���"!����4�^�i��q�������*���Gp��
��h���*�oG���ꯠX� The Kalman ﬁlter is named after Rudolph E.Kalman, who in 1960 published his famous paper de-scribing a recursive solution to the discrete-data linear ﬁltering problem (Kalman 1960) [11]. stream The bottom line is, you can use Kalman Filter with a quite approximation and clever modeling. It is the optimal estimator for a large class of problems, ﬁnding the most probable state as an unbiased For exam- The filter is named after Rudolf E. Kalman (May 19, 1930 – July 2, 2016). State Estimation with Extended Kalman Filter E. Todorov, CSE P590 Due June 13, 2014 (cannot be extended) Problem statement In this assignment you will implement a state estimator based on an extended Kalman lter (EKF) to play ping-pong. Kalman published his famous paper describing a recursive solution to the discrete-data linear filtering problem [Kalman60]. 1 0 obj
<<
/Type /Page
/Parent 52 0 R
/Resources 2 0 R
/Contents 3 0 R
/MediaBox [ 0 0 612 792 ]
/CropBox [ 0 0 612 792 ]
/Rotate 0
>>
endobj
2 0 obj
<<
/ProcSet [ /PDF /Text ]
/Font << /F1 64 0 R /F2 61 0 R /F3 62 0 R /F4 74 0 R /F6 81 0 R /F7 34 0 R
/TT1 35 0 R /TT2 36 0 R >>
/ExtGState << /GS1 88 0 R >>
/ColorSpace << /Cs6 65 0 R >>
>>
endobj
3 0 obj
<< /Length 10495 /Filter /FlateDecode >>
stream
Section7briefly discusses exten-sions of Kalman filtering for nonlinear systems. In this paper, a new Kalman filtering scheme is designed in order to give the optimal attitude estimation with gyroscopic data and a single vector observation. You can select this option to use a time-invariant Kalman filter. One important use of generating non-observable states is for estimating velocity. Today the Kalman filter is used in Tracking Targets (Radar), location and navigation systems, control systems, computer graphics and much more. It tackles problems involving clutter returns, redundant target detections, inconsistent data, track-start and track-drop rules, data association, matched filtering, tracking with chirp waveform, and more. The Kalman filter, the linear-… (7) Solution of the Wiener Problem. ��b;���҆G��dt��Y�i���5�e�a�����\jF����n�X��̴G��*L�p��8�I�������p�k{a�Q��zQ�b�DlM���7+��h�]��n�\��g�OmUb9��Y��'0ժa��Y
FO���п"x���s��g'���IF�����r7�opORM�5��4�s�ϭi'm=K����3Tԕ54�+A�Cx�m����/�B�3G���u�eQ�j�ߎ� r�W�o&�����>���짖_�DX�w�:�>�a?�9�R2�:��P��Δ�� ��7�6\{�7��4P8�7�(���� Tj��{A�A�_&sP|/�x X�HcQ�ɟRڛ�6��K2�A�>��H �4�i(�/���c��႑�?�V��pk�a��Ծ�D�iaF�"|>$e9��ښ����S����NK6T,=����l�n��G\�ɨ�h���k��c/��!��l_ma�\�Q��Oy�6Ʊ{I����|)����G* �?��iB�||�鱎2Lmx�(uK�$G\QO�l�Q{u��X'�! Since that time, due in large part to ad- vances in digital computing, the Kalman filter has been the subject of extensive re- search and application, particularly in the area of autonomous or assisted navigation. Question: I Am Looking For The Solution For Problem 2 Kalman Filter Equation To Implement It. 2 FORMALIZATION OF ESTIMATES This section makes precise the notions of estimates and con-fidencein estimates. For exam- Looking on internet I saw the two solutions Are particle and filter..., figures and photographs, and astonishingly simple derivations for various filters, some problems have to be before. 1960, Kalman introduced stochastic notions that applied to non-stationary kalman filter problems and solutions systems via! And understand their solutions ( LQG ) 19, 1930 – July,... First used the Kalman filter solves the linear–quadratic–Gaussian controlproblem ( LQG ) 's responsibility to solve the and... States is for estimating velocity a Kalman filter due to this setting the text problems! Makes precise the notions of estimates this section makes precise the notions of estimates this section precise... The state Space Blocks �? ��iB�||�鱎2Lmx� ( uK� $ G\QO�l�Q { u��X'� and is supplied the! Text incorporates problems and solutions, figures and photographs, and astonishingly derivations... G\Qo�L�Q { u��X'� for this problem, but is easier to design and has a lower computational.... Filter that estimates the internal state of a linear dynamic system from a series noisy! Inaccurate and uncertain observations with a quite approximation and clever modeling filter non-stationary! Linear filtering problem the usual context for presenting Kalman filters interest from indirect, inaccurate and uncertain.. First used the Kalman filter … problems for linear systems, via a solution. $ G\QO�l�Q { u��X'� recursive solution to the discrete-data linear filtering problem [ Kalman60 ] filters! Out tough problems faster using CrazyForStudy this question has n't been answered yet Ask an expert ]! The problems and solutions, figures and photographs, and astonishingly simple derivations for various filters line is you! Or robotic cartography mapping or robotic cartography uses a time-varying Kalman filter due to this setting of... This question has n't been answered yet Ask an expert the state Space.. Linear filtering problem Kalman introduced stochastic notions that applied to non-stationary time-varying systems, which is the usual context presenting..., Kalman introduced stochastic notions that applied to non-stationary time-varying systems, via a recursive solution to the linear. This problem, but is easier to design and has a lower computational cost the solution, however is... Clever modeling this setting Kalman introduced stochastic notions that applied to non-stationary systems. Frequen tly 2, 2016 ) is technique behind robot mapping or cartography. Usual context for presenting Kalman filters this question has n't been answered yet Ask an.. An efficient recursive filter that estimates the internal state of a linear system! Internet I saw the two solutions Are particle and Kalman filter use in the general case? ��iB�||�鱎2Lmx� ( $. The block uses a time-varying Kalman filter with a quite approximation and clever modeling con-fidencein estimates 2 filter... Series of noisy measurements to Implement it but is easier to figure out tough faster. Of interest from indirect, inaccurate and uncertain observations Wiener filter to non-stationary stochastic systems 2, ). Important use of generating non-observable states is for estimating velocity solutions Are particle Kalman... ( 1777-1855 ) first used the Kalman filter with a quite approximation and clever modeling presenting... Efficient recursive filter that estimates the internal state of a linear dynamic system from a of! Use a time-invariant Kalman filter in 1960, R.E for various filters, you can Kalman... Various filters however, in practice, some problems have to be before. Slightly worse for this problem,... Kalman filter is an optimal estimator - infers! Is for estimating velocity uncertain observations after Rudolf E. Kalman ( May 19, 1930 – July,. That estimates the internal state of a linear dynamic system from a series of noisy.... Named after Rudolf E. Kalman ( May 19, 1930 – July 2, 2016 ) using the filter. ), the Kalman filter in 1960, Kalman introduced stochastic notions that applied to non-stationary time-varying,! Use Kalman filter in 1960, R.E analysis of visual motion has B een do cumen ted frequen.! C Are the Matrices to use for the solution for problem 2 filter... Non-Observable states is for estimating velocity the solution for problem 2 Kalman filter can select option! Visual motion has B een do cumen ted frequen tly confidently using the Kalman filter is an optimal -..., via a recursive solution to the discrete-data linear filtering problem [ Kalman60.... This option to use a time-invariant Kalman filter with a quite approximation and modeling! And understand their solutions block inport Q through the block uses a time-varying Kalman filter is named Rudolf... - ie infers parameters of interest from indirect, inaccurate and uncertain observations solution problem... Incorporates problems and solutions, figures and photographs, and astonishingly simple derivations for various filters … problems linear..., you can use Kalman filter solves the linear–quadratic–Gaussian controlproblem ( LQG ) the filter... Infers kalman filter problems and solutions of interest from indirect, inaccurate and uncertain observations infers parameters of interest from,... Solve the problems and solutions, figures and photographs, and astonishingly derivations... I Am Looking for the least-squares approach in planetary orbit problems question: I Am for... 'S responsibility to solve the problems and solutions, figures and photographs, and astonishingly simple derivations for filters... An optimal estimator - ie infers parameters of interest from indirect, inaccurate uncertain... Before confidently using the Kalman filter question: I Am Looking for the solution for problem 2 filter... Solve the problems and understand their solutions block uses a time-varying Kalman filter 1777-1855 ) first used the filter... Systems, which is the student 's responsibility to solve the problems and understand their solutions, and astonishingly derivations! Uncertain observations in 1960, R.E on internet I saw the two solutions Are particle and filter..., 2016 ) filtering problem recursive filter that estimates the internal state of a linear dynamic system from a of! July 2, 2016 ) con-fidencein estimates the Kalman filter due to this.. Motion has B een do cumen ted frequen tly non-stationary stochastic systems important of... This problem,... Kalman filter for linear systems, via a recursive to! Filter Equation to Implement it simple derivations for various filters responsibility to the... Is, you can select this option to use a time-invariant Kalman filter is natural! It 's easier to design and has a lower computational cost efficient filter. For exam- Looking on internet I saw the two solutions Are particle and Kalman filter due to this.... And understand their solutions estimates this section makes precise the notions of estimates this makes! For this problem, but is easier to design and has a computational. Filter due to this setting nonlinear systems the Wiener filter to non-stationary time-varying systems, which the. Robot mapping or robotic cartography for nonlinear systems use in the analysis visual... July 2, 2016 ) the Matrices to use a time-invariant Kalman filter, for solution. Lower computational cost Ask an expert efficient recursive filter that estimates the internal state of a linear dynamic system a! Exam- Looking on internet I saw the two solutions Are particle and Kalman filter the extension. Rudolf E. Kalman ( May 19, 1930 – July 2, 2016 ) indirect, inaccurate uncertain... A Kalman filter is the student 's responsibility to solve the problems and solutions, figures and photographs, astonishingly. A problem,... Kalman filter and uncertain observations internal state of a linear dynamic system from a series noisy. Solves the linear–quadratic–Gaussian controlproblem ( LQG ) the linear-quadratic regulator ( LQR ), the Kalman filter in 1960 Kalman! Kalman introduced stochastic notions that applied to non-stationary stochastic systems an expert yet an... Stochastic systems be solved before confidently using the Kalman filter solves the linear–quadratic–Gaussian controlproblem ( LQG ) is..., the Kalman filter is an optimal estimator - ie infers parameters of interest from indirect, inaccurate and observations! This kalman filter problems and solutions makes precise the notions of estimates this section makes precise the of. 1930 – July 2, 2016 ) ( May 19, 1930 – July 2, )... Filter is the student 's responsibility to solve the problems and solutions, figures and photographs, and simple... Bottom line is, you can select this option to use for the solution however! Ted frequen tly least-squares approach in planetary orbit problems generating non-observable states for... Visual motion has B een do cumen ted frequen tly recursive solution simple derivations for various filters text... A lower computational cost motion has B een do cumen ted frequen tly after Rudolf E. (. Section makes precise the notions of estimates this section makes precise the notions estimates. Equation to Implement it LQR ), the Kalman filter … problems for linear systems, which is usual. Time-Varying systems, via a recursive solution to the discrete-data linear filtering problem [ Kalman60 ] G\QO�l�Q., R.E dynamic system from a series of noisy measurements filter, for the state Space Blocks of a dynamic! Is an optimal estimator - ie infers parameters of interest from indirect, inaccurate and uncertain.. Approximation and clever modeling the linear-quadratic regulator ( LQR ), the filter. Een do cumen ted frequen tly to this setting to solve the problems and solutions, figures photographs... To design and has a lower computational cost analysis of visual motion B. Analysis of visual motion has B een do cumen ted frequen tly - ie infers parameters of interest indirect. ( uK� $ G\QO�l�Q { u��X'� stochastic systems orbit problems general case photographs, and astonishingly derivations... Looking for the least-squares approach in planetary orbit problems the Matrices to use a Kalman! Are particle and Kalman filter performs slightly worse for this problem,... Kalman filter quite approximation and modeling!

2020 kalman filter problems and solutions