Prognoserechnung (German Edition)
Forni , M.
Privacy-preserving condition-based forecasting using machine learning
Hallin , M. Reichlin The generalized dynamic factor model, one-sided estimation, and forecasting. Journal of the American Statistical Association , — Frisch , R. Glover , K. Willems Parameterizations of linear dynamic systems: Canonical forms and identifiability. Granger , C. Princeton Studies in Mathematical Economics. Princeton University Press. Hannan , E. Econometrica 39 , — Kavalieris Multivariate linear time series models. Advances in Applied Probability 16 , — Biometrika 69 , 81 — Hazewinkel , M. Kalman Invariants, canonical forms, and moduli for linear constant finite-dimensional dynamical systems.
Mitter eds. Johnston , J. Kabaila , P. Journal of Applied Probability 20 , — Kailath , T.
Kalman , R. Cesari eds.
Services - - TED Tenders Electronic Daily
Larimore , W. Dorato eds. McKelvey , T. Helmersson System identification using an overparametrized model class—improving the optimization algorithm. Annals of Statistics 11 , — Metrika 32 , — Journal of Multivariate Analysis 21 , 29 — Econometric Theory 7 , — Corrections: Econometric Theory 10, Reinsel , G. Robinson , E. Rozanov , Y. Stock , J.
Watson Macroeconomic forecasting using diffusion indices. Willems , J. Automatica 22 , — Automatica 23 , 87 — Recommend this journal.
Econometric Theory. Who would you like to send this to? Optional message.
Prognose bei unregelmäßigem Bedarf
Altmetric attention score. All In general, the disadvantage of current weather services is that that these, if any, only indicate a rain probability. They do not disconnect strong and weak rainfall.
A statement of e. There People on the move for reliable, short-term, valid for a specific region and current forecasts of expected precipitation, e. Short term is called here in the range of a few hours and currently means that the forecasts are less than an hour old, if possible under thirty Minutes. The one by one Weather radar stations will be upcoming precipitation radar data hereby in a radar data summary calculator prepared to precipitation data that to each point in geographical Width and length, i.
Out These quantized values of precipitation intensity are given in a prediction calculator using estimation algorithms one for one or several specified future ones Timing valid prognosis for the precipitation intensity and the probability of precipitation by model extrapolation is calculated. The results of this forecast calculation will then be in the form of precalculated precipitation data to a forecast database memory issued and stored or held there.
The access of the wireless terminals to the pre-calculated precipitation data stored in the forecast database memory and thus the service control are performed via a service computer, which thus determines the interaction between the prediction database memory and the users of the terminals Interaction between the forecast data bank memory and a responsible for a terminal network location server and an individual presentation of the service controls. Of the The invention is based on the object in a method for regional Precipitation forecast locally currently predicted rainfall forecasting data for the interaction between a user and his carried by him wireless device in easy and quickly comprehensible presentation form to offer so that no meteorological knowledge of the end user are required and the local expected precipitation conditions be reproduced largely realistically.
According to the invention, which relates to a method of the type mentioned, is the above object solved by an output form the terminals, which contains a single number or an icon, picture or pictogram and in both the precipitation probability Pn as well the expected precipitation intensity I by a multiplication the value of the precipitation probability Pn with that the expected precipitation intensity I or by a nonlinear Function of the values of the precipitation probability Pn and the expected precipitation intensity I embodies are.
But also applications in which the user, i. By the invention working Regional precipitation forecasting methodology is local and temporally higher Forecasting accuracy because current data is used. Furthermore will have a higher forecasting certainty due to the use of parameter estimation and model extrapolation reached. One Advantage of the invention is also to see in it that no geographical and meteorological knowledge of the end user are necessary in the information, as well as otherwise assume it when viewing animated radar maps are.
It It is advantageous that the services computer also control of Personalization operations such as Welcome, Advertising, Billing, preselection of places or the like take over can. It is advantageous in the invention but also that a personalization in terms of preferred location s. A Sending warning messages is possible if in one area Precipitation or even heavy rainfall with high probability is to be expected. If this threshold is over or may be a separate notification of the terminal user take place, e. At the Method according to the invention is also the possibility of connection with Local servers for locating terminal networks, in particular mobile networks, an advantageous aspect.
A advantageous development of the invention is characterized by that the forecast calculator except the values quantized in the radar data summary calculator Precipitation intensity still results of further meteorological prediction models for Predictive processing can be entered what possibly the local Prediction accuracy still increases. The estimation parameters may preferably the two-dimensional motion in x longitude and y latitude directions affect.
The Location-based output of the service computer may be e. Such a route may be, for example, by a request of the user at the service computer over mobile terminal in the form "I. Will it on the way Rain? This is done by applying the above calculations for a finite Number of places along the route. So lets For example, control the hood of a parked convertible, if there is a service communicating according to the invention Car IT system is housed. Likewise, the automatic closing of windows could be in buildings Taxes.
The inventive system is characterized by the possibility the connection with local servers of mobile networks. The Invention will be described below with the aid of drawings an embodiment and two parameter estimation examples explained. In the drawings show:. According to the in 1 exemplary embodiments of the system are used in a variety of weather radar stations covering a larger area, and of which there are only four weather radar stations 1. The in the individual weather radar stations 1. For an even more exact prognosis calculation also additional results can be added meteorological predictive models are used, as in 1 by an additional prediction result block 8th symbolically indicated.
The in the forecast unit 7 precalculated precipitation data is then stored in a database or file or as in 1 in a forecast database store 9 of the forecast calculator 6 maintained. A service calculator 10 which is physically either in the radar data summary calculator 5 , in the forecast calculator 6 or as in the in 1 shown manner can be implemented in a separate physical computer, takes over the function of a service control 11 , In the illustrated embodiment controls the service control 11 the user interaction 12 the forecast database store 9 with a mobile network 13 to which a variety of wireless mobile terminals 14 belonging to end users 15 be used as a means of communication.
In the embodiment of 1 is on the service computer 10 also a personalization interaction of the service control 11 with a personalization database 16 provided so that greetings, advertisements, billing, preselections of places, routes, warning thresholds, etc. In the in 1 illustrated embodiment controls the service control 11 in the service computer 10 also the interaction 12 with a location database 17 as well as with a local server 18 a mobile radio system. In the first phase, the parameter estimator calculates 19 using a current and one or more past records 21 from the radar ten summary calculator 5 1 entered precipitation data at times t, t-1, t-2, Im in 2 More specifically, the example shown is an estimate of the parameters of the x longitude and y latitude directions, with the third dimension still having an area growth and shrinkage estimate 23 is involved.
The new and the old time delayed estimation parameters 24 respectively. In 3 is an example of the two-dimensional parameter estimation in x longitude and y latitude direction is set. In the upper row, in an x, y coordinate system, the results of two output data sets of precipitation data for the times t-1 and t are graphically in the form of four precipitation areas schematically depicted as black octagons On the far right, in the upper row, is the estimated motion vector of the precipitation areas You can see there the direction of the precipitation areas