FR2749684A1 - Data processing system for data collected from several separate monitoring stations - Google Patents

Abstract

The present invention relates to a system, in particular a computer system, for the spatial and / or temporal interpolation of a set of data. A system according to the invention comprises a plurality of measuring stations or sensors distributed at the relevant points of a geographical area, means for transmitting the data acquired by the stations or the sensors to a computer system having relevant information on them. factors influencing the spatial evolution as a function of time of the acquired values, information concerning the spatial coordinates of each sensor as well as, preferably, the date on which the various data acquisitions were made. The present invention applies to the cartographic representation of one or more values measured at a limited number of points, to an alert system in the event that a threshold value is exceeded, in particular a pollution threshold at a given location. , to the measurement of air pollution, to the production of maps and temporal animations illustrating, in particular in false colors, and / or in three dimensions, the evolution of pollution over time in a given region, to the production of chemical, biological or radioactive soil contamination maps, and epidemiology.

Description

SPATIAL AND / OR TEMPORAL INTERPOLATION SYSTEM
The present invention relates to a computer system in particular for spatial and / or temporal interpolation of a set of data.

 In many fields, we have point measurements of a value at various points in space without knowing how to draw a geographic representation of this value at points in space other than those where we made an actual measurement. .

 It is therefore an object of the present invention to provide a computer system for generating a spatial and / or temporal representation model from data measured at a limited number of points.

 It is also an object of the present invention to provide a system performing the validation of such a system.

 It is also an object of the present invention to provide a method for optimizing such a system, in particular by reducing the redundant measurement stations.

 It is also an object of the present invention to provide a system for automatically generating distribution maps, advantageously continuous, of a value or of several values measured at a limited number of points.

 These aims are advantageously achieved by a system according to the invention comprising a plurality of measurement stations or sensors distributed at the relevant points of a geographical area, means for transmitting the data acquired by the stations or the sensors to a computer system having relevant information on the factors influencing the spatial evolution as a function of time of the acquired values, information concerning the spatial coordinates of each sensor as well as, preferably, the date on which the various data acquisitions were carried out.

 The main object of the invention is a system for drawing up a spatial distribution map of the value of at least one characteristic in an area, characterized in that it comprises a plurality of measurement stations distributed in the area, means, transmissions of the data acquired by the various stations to calculation means drawing up from the spatial distribution of the values measured at a limited number of points, a map of the distribution of these values over the entire area to be mapped.

 The invention also relates to a system, characterized in that it further comprises means for synchronizing the measurement stations so that all the measurement stations simultaneously carry out the measurements of the value of the characteristic monitored in the area to be mapped.

 The invention also relates to a system, characterized in that it performs an interpolation of the values measured by the various measurement stations.

 The invention also relates to a system, characterized in that the interpolation uses the Kriging algorithm.

 The invention also relates to a system, characterized in that the calculation means carry out corrections or weightings of estimated values of the characteristic at various points of the area to be mapped as a function of the additional information concerning the factors influencing the spatial diffusion. of the monitored characteristic.

 The invention also relates to a system, characterized in that it further comprises a mobile measurement station.

 The invention also relates to a system, characterized in that the system is a system for mapping air pollution in a geographic area.

 The subject of the invention is also a system, characterized in that it includes means for displaying colors and in that the means for calculating produce a map in false colors of the values of the characteristic monitored in the area to be mapped.

 The invention also relates to an alert system, characterized in that it comprises a system for drawing up a spatial distribution map of the value of at least one characteristic in an area according to the invention detecting whether the concentration of at least one harmful product does not locally or globally exceed a predetermined threshold, connected to an alarm device triggered by the system for drawing up a spatial distribution map of the value of at least one characteristic in a zone, when the predetermined threshold is reached or exceeded.

 The invention also relates to a method for optimizing a system, characterized in that it comprises a step consisting in identifying any redundant measurement stations and a step of eliminating identified redundant stations.

The invention also relates to a method, characterized in that it comprises the steps consisting in
a) draw up a map of the area to be mapped from the values measured by all the measurement stations
b) develop a map from the measurements made by a subset of the measurement stations
c) compare the map prepared from the values provided by all the stations (a) with that drawn from the data provided by the subset of the stations (b); and
d) in the case where the maps produced in stages a) and b) are similar, eliminate the stations which do not belong to the subsets of the stations which were used to prepare the map in stage b).

The invention will be better understood by means of the description below and the appended figures given as nonlimiting examples and in which
- Figure 1 is an explanatory diagram of the operation of the system according to the present invention
- Figure 2 is a diagram illustrating the correspondence of the various types of data capable of being processed by the system of Figure 1;
- Figure 3 is a flow diagram illustrating the optimization of the system of Figure 1.

 In FIGS. 1 to 3, the same references have been used to designate the same elements.

 In FIG. 1, we can see a device according to the present invention comprising a plurality of data acquisition stations 1.1 to 1.N connected by communication lines 3 and / or 3.1 to 3.N to calculation means, in particular to a computer 5. The calculation means 5 of the system according to the invention comprise for example a microcomputer, a graphics station, a mini-computer or a scientific super-computer, depending on the number and type of calculation to be carried out and the desired response time. An example of the geographic distribution of acquisition stations is illustrated in FIG. 2. In addition to the fixed stations 1.1 to 1.7, the example illustrated also comprises a mobile station i a. The information storage means of the computer 5 include information on the location of each acquisition station, for example in the form of two or three dimensional spatial coordinates. The coordinates of the acquisition stations 1.1 to 1.N are symbolized in FIG. 2 by a layer 7 superimposed on a cartographic layer 9 and on a layer 11 containing information relating to factors liable to influence the diffusion of the measured characteristics. The data acquisition stations 1.1 to 1.N measure, for example, air pollution and in particular the gas concentration NO, NO2, CO, C02, SO2; O3 being understood that the measurement of other data, such as for example the concentration of dust or pollutants of the atmosphere in general, in particular in radioactive gases or harmful chemical or biological products, the measurement of temperature, direction and / or wind speed, concentration of products in the soil, in particular heavy metals such as lead, cadmium or nickel, the measurement of which is particularly advantageous near polluting factories, population census, diseases by sector, and more generally of any data for which it is desired to know the spatial distribution and / or the evolution over time, does not depart from the scope of the present invention. Each station is advantageously provided with an automatic call device on switched telephone networks or with a dedicated transmission line. Advantageously, the data are digitized and transmitted by the networks in digital form, for example via a modem. The layer 9 corresponds to the geographic data advantageously stored in digital form in the computer 5. The data of the layer 9 comprises for example the information concerning the natural reliefs, the buildings, the rivers 13, and / or the administrative limits 15. The layer 11 may comprise, on the one hand, the data which do not change over time, such as for example the relief influencing the propagation of the pollution, the nature of the soils, the presence of updrafts or stationary horizontal flows. or periodic in the very low layers of the atmosphere, the nature of the soil, etc. This type of data is permanently stored in the mass memory of the computer 5. This data can be supplemented by variable data illustrated by the arrow 17 in FIG. 1 concerning for example the meteorological conditions, in particular the temperature in various points of the region concerned, wind speed and direction and / or precipitation, information from the analysis of satellite images or information concerning the intensity of traffic in the area concerned. The information 17 includes a validity date corresponding to the moment of acquisition, in the case of real data, or the instant of an evolution in the case of a forecast or a simulation (weather forecast or traffic forecast ). Likewise, the information 17 can come from an internal database of the computer 5 concerning previously treated cases and / or cards previously produced by the system according to the present invention.

 In a first alternative embodiment corresponding to the case where there is a very large computing power and / or the number of data or the complexity and the extent of the territory to be mapped are low, the calculations of the spatial evolution of values measured over time. For example, the laws of diffusion, the equation of heat, the laws of themodynamics and fluid mechanics or the so-called Monte-Carlo statistical laws are used. The calculation data can be corrected or weighted by the data 17. For example, a very long building can constitute a barrier that cannot be easily crossed for pollutants. Pollutants move faster on smooth soil than on rough soil while an updraft decreases the concentration of pollutants at ground level. In a second variant embodiment, the system according to the present invention performs statistical calculations on the diffusion of the measured values. For example, a calibration of the system is carried out by means of complementary measurements carried out with the mobile station 1.a, the calibration provides a heuristic statistical base on values at various points as a function of the values measured at the level of stations 1.1 to 1. N, as well as data 17. For example, the data 17 may include information relating to automobile traffic at various points in the geographical area to be monitored and mapped.

 In a third alternative embodiment, an interpolation of the data from the various stations 1.1 to 1.N is corrected and weighted by the complementary values coming from the layer 11 or from the data 17. In the preferred embodiment, the so-called Kriging method with different variogram models, including spherical, linear, exponential, Gaussian or others. Advantageously, the most suitable variogram is continuously tested during calculations so as to choose the most efficient.

 Kriging is an objective analysis method based on the theory of random functions. It consists in forming a linear estimator which is without bias even if the field of normals has a drift, and optimal, in the sense that it minimizes the mean square error. This method allows optimal spatial interpolation of the data by minimizing the error and the variance of the error between the true values and the estimated values.

Kriging is described in particular in Matheron, 1963,
Treaty of Applied Geostatistics, volume II, Le Krigeage, Memoirs of BRGM n "14, Ed of BRGM, 172 p, and in Isaaks and Srivastava, 1989, Applied
Geostatistics, Oxford University Press, New York, 561 p.

 Other interpolation methods are possible, among them we will choose for example the method of least squares and that of spline functions with minimum curvature.

 The least squares method consists in fitting a field, either locally or globally, to a linear combination of basic functions chosen in advance, by minimizing the sum of the quadratic deviations at the experimental points. For an interpolation, we generally limit ourselves to performing a polynomial regression to a quadric or a cubic.

 The method of indeterminate coefficients consists in adapting a polynomial function (the degree of which increases with the number of points) which coincides with the measured values.

 The triangulation method creates a patchwork of triangles whose sides pass through the measurement points. From these triangles, for each point of the mesh, the estimate is made by performing a linear interpolation, for example.

 The distance weighting method: the weights are decreasing functions of the distances (in general the inverse of the square of the distances).

 Shepard's method is based on the previous method of inverse distances weighted by the method of least squares.

 The method of spline functions with minimum curvature is used for smoothing and determined by minimization of a function which must pass closest to the points and constitute a surface of minimum curvature. With the spline functions of the "thin plate" type, we get closer to the characteristics of Kriging.

 Corrections are made, for example, based on knowledge of dynamics, thermodynamics, the lower layers of the atmosphere.

 As a variant, the program of spatial interpolation calculations can also have a base of interpolation method which is continuously tested with the production of an error map to choose the most appropriate method as a function of the numerical values to be processed.

 Advantageously, the pollution mapping system includes a function allowing, after the designation of a route to be carried out in the monitored geographical area, for example the Paris region, to determine the quantity of pollutants absorbed by an individual during a route . The speed of the route is estimated based on traffic conditions and / or the distance. The integration over time of the inhalation of pollutants makes it possible to calculate the total quantity of pollutants inhaled during the journey. It is thus possible to determine the correlation between various diseases and the inhalation of pollutants or determine optimized routes to reduce the quantity of pollutants absorbed.

 A first variant embodiment of the system according to the present invention uses commercial software. For example, a database management system or a specific program written for example in Visual Basic (marketed by the company MICROSOFT) ensures the launch of scientific calculation software, such as for example, the software marketed under the name Matlab , ensuring the interpolation calculations, the statistical calculations or the diffusion calculations, the fluid mechanics calculations or others to be carried out. The fusion of interpolated data with a digital geographic map, for example by the software marketed under the name Map info, is ensured by the software marketed under the name Surfer. As a variant, the program merges the interpolated data with a digital geographic map. The geographic map in question must be digitized beforehand and digitized by appropriate software, for example from the National Geographic Institute. The resulting map is displayed in the Matlab software or advantageously in the presentation software, such as for example marketed under the name Powerpoint, which can make an animation, advantageously in false colors, of the evolution of the pollution in the geographical area treated with the time. Pollution cartographic images can be produced in two dimensions or in three dimensions.

 In a second variant embodiment, the system according to the present invention was previously written in a computer language interpreted and / or compiled. Excellent results have been obtained using the software sold under the name Matlab, in particular with regard to the functions allowing interpolation and graphic representations (maps and animations).

 The choice of the location of acquisition stations 1.1 to 1.N is decisive in the efficiency of the system according to the present invention. In the first case, the existing stations must be used. In such a case, the model is advantageously validated by the use of a mobile station 1.a making it possible to compare the calculation results with an actual measurement at various points in the area to be mapped. As a variant, the stations are distributed regularly and / or randomly over the area to be mapped.

Then, advantageously, the unnecessary or redundant stations are deleted or moved. An example of a method for eliminating unnecessary and / or redundant stations is illustrated in FIG. 3.

 The program begins in 21.

 We go to 23.

 In 23, the data from stations 1.1 to 1.N are acquired.

 We're going to 25.

 At 25, a model M of the references corresponding to the use of all the stations 1.1 to 1.N is calculated.

 We're going to 27.

 At 27, a counter is initialized.

 We are going to 29.

 At 29, the counter is incremented.

 We're going to 31.

 At 31, a model Mi is calculated, based on the data of all the stations with the exception of the station i examined (designated by the counter 29).

 We're going to 33.

 In 33, we compare the result of the calculation elaborated from the data of all the stations (in 25) and that elaborated without taking account of the station i (in 31). The comparison 33 takes account of the error threshold which one wishes to tolerate.

 If the models are different, we go to 35.

 If the models are substantially equal, we go to 37.

 In 37, we delete the station i from the list of stations whose data are examined for the development of the model M (in 25).

 In 35, we check if we have reached the last station. If not, we go to 29.

 If yes, we go to 39.

 In 39, the optimization of the network of acquisition stations is completed.

 In a variant of the method of FIG. 3, the essential stations are firstly determined by the loop comprising steps 29 to 35, the deletion of which disturbs the measurement results, then the models are examined not by deleting the stations one by one, but per pair of stations, triplet of stations, quadruple of stations, etc.

until validation of the deletion of the maximum number of redundant or unnecessary stations. Redundant stations are deleted and / or moved to a non-redundant location. As a variant, before the stations are installed, one or more mobile stations are moved extremely rapidly between various points so as to ensure the acquisition step 23 almost instantaneously (almost simultaneous measurements at various points), then the places are determined. it is essential to have the stations when building the system according to the present invention.

 Similarly, it is possible to apply the disturbance method to move the various stations to an optimal location. Limited depications of a station are carried out in various directions and the results obtained are compared with the results measured by the mobile station at various points so as to retain the most favorable location for placing a station. The disturbance is applied successively to some or advantageously to all stations until the optimal locations of the various measurement stations 1.1 to 1.N are determined.

 The mobile station 1.a validates a model. In fact, the station is moved to various points and the pollution rate at this measured point is compared with the pollution rate predicted by the system. In case of divergence, the algorithm used, the layout or the number of stations are modified.

 The present invention applies to the cartographic representation of one or more values measured at a limited number of points, to an alert system in the event that a threshold value is exceeded, in particular a pollution threshold at a given location. , measuring air pollution, producing maps and time animations illustrating, in particular in false colors, and / or in three dimensions, the evolution of pollution over time in a given region, the creation of maps of chemical, biological or radioactive contamination of the soil, and of epidemiology.

 In other applications of the present invention, two geographical representations are made (maps and / or animations) illustrating the temperature field, the wind field, the population density, the density of the diseases by sector, and more generally of any parameter whose spatial distribution and / or monitoring over time is desired.

Claims (11)

 1. System for developing a spatial distribution map of the value of at least one characteristic in an area, characterized in that it includes a plurality of measurement stations (1.1 to 1.N) distributed in the area , means (3, 3.1 to 3.N), transmissions of the data acquired by the various stations to calculation means (5) elaborating from the spatial distribution of the values measured at a limited number of points, a map of the distribution of these values over the entire area to be mapped.  2. System according to claim 1, characterized in that it further comprises means for synchronizing the measurement stations (1.1 to 1.N) so that all the measurement stations (1.1 to 1.N) perform simultaneously the measurements of the value of the characteristic monitored in the area to be mapped.  3. System according to claim 2, characterized in that it performs an interpolation of the values measured by the various measurement stations (1.1 to 1.N).  4. System according to claim 3, characterized in that the interpolation uses the Kriging algorithm.  5. System according to any one of the preceding claims, characterized in that the calculation means (5) carry out corrections or weightings of estimated values of the characteristic at various points in the area to be mapped as a function of the additional information (17 ) concerning the factors influencing the spatial diffusion of the monitored characteristic.  6. System according to any one of the preceding claims, characterized in that it further comprises a mobile measuring station (1.a).  7. System according to any one of the preceding claims, characterized in that the system is a system for mapping air pollution in a geographical area.  8. System according to any one of the preceding claims, characterized in that it includes color display means and in that the calculation means (5) draw up a map in false colors of the values of the characteristic monitored in the area to be mapped.  9. Alert system, characterized in that it comprises a system for drawing up a spatial distribution map of the value of at least one characteristic in an area, according to any one of the preceding claims detecting whether the concentration in at least one harmful product does not locally or globally exceed a predetermined threshold, connected to an alarm device triggered by the system for drawing up a spatial distribution map of the value of at least one characteristic in an area , when the predetermined threshold is reached or exceeded.  10. A method of optimizing a system according to any one of the preceding claims, characterized in that it comprises a step consisting in identifying any redundant measurement stations (1.1 to 1.N) and a step of elimination. of redundant stations identified.  11. Method according to claim 10, characterized in that it comprises the steps consisting in  a) draw up a map of the area to be mapped from the values measured by all the measurement stations (1.1 to 1.N)  b) develop a map from the measurements made by a subset of the measurement stations (1.1 to 1.N)  c) compare the map prepared from the values provided by all the stations (a) with that drawn from the data provided by the subset of the stations (b); and  d) in the case where the maps produced in stages a) and b) are similar, eliminate the stations which do not belong to the subsets of the stations which were used to prepare the map in stage b).