JP2013065208A - Substance discharge estimation device, method for the same and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable substance discharge discharged from a discharge point to be estimated and to improve accuracy of diffusion projection based on the discharge.SOLUTION: A substance discharge estimation device 1 comprises: an information acquisition section 21 which acquires positional information of each observation point, observation value of concentration observed at the observation points as well as observation times and also acquires positional information of a discharge point as well as a substance discharge start time; an influence coefficient calculation section 22 which calculates an influence coefficient determined, by using a diffusional model, as a function of a relative position between a position of a particle and the discharge point as well as relative time between a substance discharge time and the observation time at the observation point; an estimation section 24 which estimates discharge strength of the particle discharged from the discharge point by minimizing a residual norm represented by using differences between observation values of concentration at the plurality of the observation points and calculated values of the concentration at each observation point obtained by carrying out operations; and a discharge strength correction section 25 which corrects the discharge strength of each particle estimated by the estimation section 24 through weighting in accordance with the influence coefficient.

Description

  The present invention relates to a substance release amount estimation apparatus, a method thereof, and a program for estimating a release amount released from a specific release position.

  In response to the release of pollutants due to accidents at plant facilities (thermal power plants, garbage incineration facilities, chemical plants, etc.) or the release of toxic gases, etc. due to terrorism, etc. Techniques have been proposed for estimating source information (position of release point and substance release amount).

  For example, Non-Patent Document 1 evaluates the influence at an observation position based on information such as a virtual release point and time, obtains a release point that minimizes an error by a variational principle, and determines the time and substance at the release point. A method for estimating the release amount has been proposed.

Yoshihiro Ishida, Shinsuke Kato, Yusuke Hatakeyama, “Estimation method of environmental impact substances based on response coefficient method using sensing information (2nd report)”, Proceedings of the 2009 Air Conditioning and Sanitary Engineering Conference (September 2009)

  By the way, in the case of an accident at a plant facility or the like, there are many cases where the release point and release time of the pollutant are clear. In this case, the process of estimating the release point becomes unnecessary, and the process can be simplified. .

  The present invention has been made in view of such circumstances. When the release point and release time of a substance are clear, it is possible to estimate the release amount of the substance released from the release point. It is an object of the present invention to provide a substance release amount estimation device, method and program for improving the accuracy of diffusion prediction based on the release amount.

In order to solve the above problems, the present invention employs the following means.
The present invention relates to a substance release amount estimation device for estimating the amount of substance released from a specific release point based on concentration observation values observed at a plurality of observation points, and the position of each observation point The information, the concentration observation value observed at the observation point and the information of the observation time, the information acquisition means for acquiring the position information of the release point and the release start time of the substance, and the diffusion model Coefficient calculating means for calculating an influence coefficient determined in accordance with the relative position between the emission point and the release time and the relative time between the substance release time and the observation time at the observation point; Estimate the emission intensity of particles emitted from the emission point by minimizing the residual norm expressed using the difference from the calculated concentration at each observation point An estimation means for estimating a release amount from the release point using the estimated emission intensity of the particles, and an influence coefficient calculated by the influence coefficient calculation means for the emission intensity of each particle estimated by the estimation means There is provided a substance emission amount estimating device comprising emission intensity correcting means for correcting by weighting according to the above.

According to the present invention, the information acquisition means acquires the position information of a plurality of observation points, the concentration observation values and the observation time there, and the substance release point and the release start time. The influence coefficient calculation means calculates an influence coefficient determined according to the relative position between the position of the particle and the release point and the relative time between the substance release time and the observation time at the observation point, using the diffusion model. The estimation means minimizes the residual norm expressed by using the difference between the observed concentration values at multiple observation points and the calculated concentration value at each observation point, thereby calculating the particle emission from the emission point. The emission intensity is estimated, and the emission amount from the emission point is estimated using the estimated emission intensity of the particles. In the emission intensity correcting means, the emission intensity of each particle estimated by the estimating means is corrected by weighting according to the influence coefficient calculated by the influence coefficient calculating means.
In this way, the emission intensity of each particle estimated by the estimation means is corrected by a weight determined according to the degree of influence each particle has on the observation point, so that a large number of particles released at the same time It is possible to make a difference in strength. As a result, the identification accuracy at each observation point can be increased, and the accuracy of the diffusion prediction in the subsequent processing performed using this identification result can be increased.

  In the release amount estimation device for the substance, the release intensity correction means is configured so that the release quantity obtained from the corrected release intensity of the particles matches the release quantity estimated by the estimation means. The emission intensity may be corrected.

  In this way, since only the emission intensity of each particle is corrected without changing the value of the emission amount estimated by the estimation means, the identification accuracy can be further improved.

  The present invention is a substance release amount estimation method for estimating the release amount of a substance released from a specific release point based on concentration observation values observed at a plurality of observation points, the position of each observation point being Information, the concentration observation value observed at the observation point and the information of the observation time, the information acquisition process for acquiring the position information of the release point and the release start time of the substance, and the diffusion model Coefficient calculation process for calculating the influence coefficient determined according to the relative position between the release point and the release time of the substance and the relative time between the substance release time and the observation time at the observation point; Estimate the emission intensity of particles emitted from the emission point by minimizing the residual norm expressed using the difference from the calculated concentration at each observation point An estimation process for estimating the release amount from the release point using the estimated emission intensity of the particles, and an influence coefficient calculated in the influence coefficient calculation process for the emission intensity of each particle estimated in the estimation process There is provided a method for estimating a release amount of a substance having a release intensity correction process that is corrected by weighting according to.

  The present invention is a substance release amount estimation program for estimating a release amount of a substance released from a specific release point on the basis of concentration observation values observed at a plurality of observation points. The position information, the concentration observation value observed at the observation point and the information of the observation time, the information acquisition process for acquiring the position information of the release point and the release start time of the substance, and the diffusion model, An influence coefficient calculation process for calculating an influence coefficient determined according to a relative position between the position of the release point and the release point, and a relative time between the substance release time and the observation time at the observation point, and concentration observation values at a plurality of the observation points And the residual norm expressed using the difference between the calculated concentration at each observation point and the calculation result is minimized, thereby releasing particles released from the release point. An estimation process for estimating the intensity and estimating the release amount from the release point using the estimated emission intensity of the particles, and calculating the emission intensity of each particle estimated in the estimation process in the influence coefficient calculation process There is provided a program for estimating a released amount of a substance for causing a computer to execute a release intensity correction process for correcting by weighting according to an influence coefficient.

  According to the present invention, it is possible to estimate the release amount of a substance released from a release point, and to improve the accuracy of diffusion prediction based on the release amount.

It is a figure for demonstrating the outline of discharge | release amount estimation implemented by the discharge | release amount estimation apparatus of the substance which concerns on one Embodiment of this invention. It is the block diagram which showed an example of the hardware constitutions of the discharge | release amount estimation apparatus of the substance which concerns on one Embodiment of this invention. It is the functional block diagram which showed the various functions with which the discharge | release amount estimation apparatus of the substance which concerns on one Embodiment of this invention is provided. It is the figure which compared and showed the condition of each particle | grain in the observation point with and without performing the correction | amendment of the discharge | release intensity | strength of a particle | grain. It is a flowchart explaining the discharge | release amount estimation method of the substance which concerns on one Embodiment of this invention.

A substance release amount estimation apparatus and method according to an embodiment of the present invention will be described below with reference to the drawings.
A substance release amount estimation device (hereinafter simply referred to as “release amount estimation device”), a method thereof, and a program according to an embodiment of the present invention include, for example, a time from a specific release point A as shown in FIG. When release of the substance is started at t ₁ , release from the release point A using the concentration observation values measured by the observation devices installed at a plurality of observation points B1 to Bn (n is an integer of 2 or more) The release intensity of the released substance (particle) and the release amount Q are identified. More specifically, in the present embodiment, continuous release in which a substance is continuously released from the release point A is assumed (instantaneous release is a part thereof), and the amount of substance released at each time in this case is identified. FIG. 1 illustrates the case where n = 5 at the observation point.

  Note that the observation points B1 to Bn are not necessarily at the same position, and may be different positions, for example. That is, it is only necessary that the position information of each observation point and the concentration observation value at that point can be acquired in association with each other.

  The release amount of the substance and the release intensity of each particle estimated by the release amount estimation apparatus are input to the substance diffusion prediction apparatus, for example, and used for the substance diffusion prediction. In the substance diffusion prediction, for example, weather data, topographic data, or the like is used, and the diffusion state of the substance released from the emission point A is predicted in time series.

FIG. 2 is a block diagram showing an example of a hardware configuration of the discharge amount estimation apparatus 1 according to an embodiment of the present invention.
As shown in FIG. 2, the discharge amount estimation device 1 includes, for example, a CPU 11, a ROM (Read Only Memory) 12 for storing a program executed by the CPU 11, and a RAM that functions as a work area when each program is executed. (Random Access Memory) 13, a hard disk drive (HDD) 14 as a mass storage device, a communication interface 15 for connecting to a network, an access unit 17 to which an external storage device 16 is attached, a keyboard, a mouse, and the like An input unit 19 and a display unit 20 including a liquid crystal display device for displaying data. These units are connected via a bus 18.

The ROM 12 is equipped with a release amount identification program for identifying the release amount of a substance released from a specific release point A, and the CPU 11 reads out this program to the RAM 13 and executes it, thereby performing various processes described later. Realized.
In the configuration example shown in FIG. 1, the storage medium for storing the program executed by the CPU 11 is not limited to the ROM 12. For example, other auxiliary storage devices such as a magnetic disk, a magneto-optical disk, and a semiconductor memory may be used.

  FIG. 3 is a functional block diagram showing various functions provided in the release amount estimation device 1. As shown in FIG. 3, the discharge amount estimation device 1 includes an information acquisition unit 21, an influence coefficient calculation unit 22, an estimation unit 24, and a discharge intensity correction unit 25.

  The information acquisition unit 21 acquires information from observation devices provided at the observation points B1 to Bn, that is, position information, concentration observation values, and measurement times of the observation points. Here, the method of obtaining information from the observation points B1 to Bn is not limited to the example of obtaining from each observation device. For example, if there is a gas concentration observation system by a private or public institution and it has a database function that sequentially accumulates concentration data at each observation point B1 to Bn, it is installed on a network such as the Internet. The obtained database may be accessed to obtain the position information, concentration observation value, and observation time of each observation point.

  Further, the information acquisition unit 21 acquires the position information of the release point A and the information of the substance release start time. For example, the position information and the release start time of the release point A may be input to the user from the input unit 19 or may be obtained from another system via the communication interface 15. .

The influence coefficient calculation unit 31 uses the various input information acquired by the information acquisition unit 21 to calculate the influence coefficient φ _ki using a diffusion model.

The influence coefficient φ _ki represents the concentration at the observation point B _k (1 ≦ k ≦ n) due to the particles emitted at the unit intensity at the time t _i , and the position (x _i , y _i , z _i ) of the particles and position at the observation point _{B k (x K, y K} , z K) relative position, and is calculated in accordance with the relative time between the measurement time _{t k} at the observation point _{B k} and substance release time _{t i.} Here, the influence coefficient φ _ki is calculated, for example, for each particle l (el) emitted at time t _i .

For example, a puff model is used as a diffusion model when assuming a flat uniform flow field (a state where the terrain of the diffusion region is flat and the wind is flowing uniformly). According to the puff model, when the wind speed is U [m / sec], the influence coefficient φ _ki is given by the following equation.

In equation (1), σ _x , σ _y , and σ _z are diffusion parameters [m] in the x, y, and z directions of the concentration distribution, respectively, and are obtained based on the Pasquill-Gifford diagram and empirical formulas. The diffusion model is not limited to the puff model, and other models such as a plume model can also be used.

  In general urban areas and the like, there are many cases where the flow is not uniform due to the influence of buildings, topography, etc. In such a complex airflow field, the influence coefficient is calculated by numerical diffusion calculation. That is, the concentration (that is, the influence coefficient) at the evaluation point when the unit intensity is discharged from the discharge point A is obtained using various simulation models. Examples of the simulation model include a modified plume model, a potential flow model, and a viscous flow model. Details of these are disclosed, for example, in Mitsubishi Heavy Industries Technical Report vol.21 No.5 reprint (September 1984) “Development of numerical simulation model for smoke diffusion” pp.1-8. In addition, using the plume / puff model detailed in the "Regulation Manual for Total Amount of Nitrogen Oxides" edited by the Air Quality Control Bureau of the Environment Agency, the well-known model, the particle method in a cell, and the Lagrangian particle model. Can also be sought.

The estimation unit 24 calculates the emission intensity q _i when the residual norm π calculated using the influence coefficient takes a minimum value, and further, the number of emitted particles at time t _i is calculated based on the emission intensity q _i. By multiplying by L _i , the release amount Q _i (= q _i × L _i ) at time t _i is calculated.
Here, q _i is the emission intensity of one particle emitted at time t _i , and it is assumed that L _i particles having the same intensity are emitted at the same time.

  The residual norm π is given by the following equation (2).

In equation (2), f _k is the concentration observation value at the observation time t _k at the observation point B _k . N is the number of observation points. Further, F _k is a concentration calculation value at the observation time t _k at the observation point B _k . For example, the concentration calculation value F _k by the particle method is given by the following equation (3).

In (3), (x _k , y _k , z _k ) is position information (x _k , y _k , z _k ) at the observation point B _k , t _k is the observation time at the observation point, and t _i is the emission. Time, q _i is the emission intensity of one particle, and L _i is the number of particles released at time t _i .

  Substituting the above equation (3) into equation (2) yields the following equation (4).

The estimation unit 24 obtains the emission intensity q _i when the residual norm π takes the minimum value by solving the simultaneous equations of the expression (5) derived from the expression (4). Then, to calculate the released amount _{Q i} of the emission point A in the release time _{t i} from the emission intensity _{q i.}

In the case of instantaneous release, the above equation (5) is expressed by the following equation (6) from φ _ki = 0 (i ≠ 1).

Here, in order for the above equation (5) to have a solution, normally M <N, that is, the number of observation points N (including time) needs to be larger than the number of time steps M at the release time. However, since the number of time steps M at the release time is usually larger than the number N of observation points, there is an inconvenience that simultaneous equations relating to q _i obtained from the above equation (5) cannot be solved.

Therefore, the number of time steps M at the release time is adjusted to be equal to or less than the number N of observation points, and the amount of discharge between time steps is approximated by linear interpolation.
That is, the emission intensity q _{t at} time t (t _i <t <t _{i + 1} ) can be obtained by the following equation (7).

  Further, when the above equation (4) is expressed using approximation by linear interpolation, the following equation (8) is obtained.

Then, similar to the above, simultaneous equations relating to q _i as shown in the following equation (9) are obtained.

  In the case of i = 1 (instantaneous emission), the relationship of the following equation (10) is satisfied in the above equation (8), which is consistent with the equation (6).

The estimation unit 24 may obtain the emission intensity q _i using the above equation (5), or may obtain the emission intensity q _i using the equation (9). Which formula is used may be determined according to the magnitude relationship between the number N of observation points and the number of time steps M, for example. In the case of instantaneous release (i = 1), the equation (6) may be used.

The emission intensity correction unit 25 corrects the emission intensity p _i of each particle obtained by the estimation unit 24.
The estimation unit 24 estimates the emission intensity q _i on the assumption that the emission intensity of each particle emitted at the same time is uniform. Actually, the emission intensity q _i of each particle with respect to the observation point B _k is estimated. Is not uniform. Therefore, the emission intensity correction unit 25, according to the degree of influence of the observation point B _k, to correct the emission intensity q _i of the individual particles, increasing the assimilation of the accuracy of the observation point B _k.

Specifically, the influence coefficient φ _ki (i = 1 to M) calculated for each particle l (el) in the influence coefficient calculation unit 22 has a maximum value q _mk , and the influence coefficient The maximum value is φ _km , and the emission intensity of each particle is corrected by weighting according to the maximum influence coefficient φ _km and the emission intensity q _mk information and the influence coefficient φ _ki of each particle. The corrected emission intensity q _i ′ is given by the following equation (11). The correction term is set so that the emission intensity increases as the influence coefficient φ _ki increases.

  Then, by reflecting the emission intensity corrected by the above expression (11) in the above expression (4), the expression (12) is derived, and from the above expression (12), the above expression (5) is converted into the following (13) It is expressed by a formula.

The accuracy of diffusion prediction can be improved by performing diffusion prediction subsequent to the emission intensity of each particle corrected by the above equations (11) and (13).
In the above equation (11), since the emission intensity of each particle is calculated by adding a weighting coefficient corresponding to the influence coefficient φ _ki to the radiation intensity q, the total intensity Qi ′ of all the particles after correction is The amount of radiation Qi calculated by the estimation unit 24 does not match.
Therefore, in order to match the amount of radiation Qi' after correction in observation point B _k to the radiation amount Qi calculated by the estimator 24, corrects the emission intensity of each particle using the following equation (14).

FIG. 4 shows a comparison of the state of each particle at the observation point _Bk when the particle emission intensity is not corrected and when the particle emission intensity is not corrected. In FIG. 4, for convenience, it is assumed that particles are continuously released one by one per second, FIG. 4 (a) shows the state of the particles when the emission intensity is not corrected, and FIG. 4 (b). The state of the particles when the emission intensity is corrected is shown. When the correction shown in FIG. 4 (a) is not performed, the concentration at the observation point B20 is given by the following equation (15), but when the correction shown in FIG. 4 (b) is performed, the observation point B20. The density Q at is given by the following equation (16), which shows that the assimilation accuracy is improved.

  Next, a method for estimating the release amount of a diffusing substance in the release amount estimation apparatus 1 having the above-described components will be described with reference to FIG. Here, FIG. 5 is a flowchart for explaining the method for estimating the released amount of the substance according to the present embodiment.

  First, in step SA1, the information acquisition unit 21 acquires position information, concentration observation values, and measurement time information at each of the observation points B1 to Bn, and acquires position information and a discharge start time of the discharge point A. .

Next, in step SA2, the influence coefficient calculation unit 32 calculates the influence coefficient φ _ki using a diffusion model (for example, Expression (1)).

Next, in step SA3, the estimation unit 24, residual norm π is calculated emission intensity q _i which minimizes, by the release number is multiplied by the emission intensity q _i, the time t _i in the release point A Release amount Q _i is estimated.

Next, in step SA4, the emission intensity of each particle identified by the estimation unit 24 is corrected. Calculation results such as the emission intensity and the emission amount Q _i of each particle after correction are input to a system that performs diffusion prediction, which is subsequent processing, and used in diffusion prediction.

As described above, according to the diffusion substance release amount estimation device 1, the release amount estimation method, and the program according to the present embodiment, the information acquisition unit 21 uses the position information of a plurality of observation points, concentration observation values and observations there. The time is acquired, and the release point and release start time of the substance are acquired. Then, the influence coefficient calculation unit 22 calculates the influence frequency at each observation point using the information obtained by the information acquisition unit 21. Subsequently, the estimation unit 24 calculates the emission intensity q _i when the residual norm π expressed using the influence frequency calculated by the influence coefficient calculation unit 22 takes the minimum value, and further the emission intensity correction unit 25. The emission intensity q _i is corrected by weighting according to the influence concentration, whereby the emission intensity for each particle is calculated.

Thus, according to the present embodiment, by minimizing the residual norm calculated from the difference between the concentration observation values at the plurality of observation points B1 to Bn and the calculated concentration value at the observation point obtained by calculation, The emission intensity of the particles emitted from the emission point A is estimated, and the emission quantity Q _i is estimated using this emission intensity. Furthermore, since the emission intensity of each particle is corrected by a weight determined according to the degree of influence on the observation point, it is possible to make a difference in the intensity of the emission intensity among a number of particles emitted at the same time. It becomes possible. Thus, since the emission intensity of the particles is individually corrected, the identification accuracy at each observation point can be increased, and the accuracy of the diffusion prediction in the subsequent processing performed using this identification result can be increased. .

1 Substance release amount estimation device 15 Communication I / F
19 Input unit 21 Information acquisition unit 22 Influence coefficient calculation unit 24 Estimation unit 25 Release intensity correction unit

Claims (4)

A substance release amount estimation device that estimates the amount of substance released from a specific release point based on concentration observation values observed at a plurality of observation points,
Information acquisition means for acquiring position information of each observation point, concentration observation value observed at the observation point and information of observation time, and acquiring position information of the release point and release start time of the substance;
Using a diffusion model, an influence coefficient calculating means for calculating an influence coefficient determined according to a relative position between the position of the particle and the emission point, and a relative time between the substance release time and the observation time at the observation point;
Release of particles emitted from the emission point by minimizing the residual norm expressed using the difference between the concentration observation value at the plurality of observation points and the calculated concentration value at each observation point obtained by calculation. Estimating means for estimating intensity and estimating the emission amount from the emission point using the estimated emission intensity of the particles;
A release amount estimating device for a substance, comprising: a release intensity correcting means for correcting the emission intensity of each particle estimated by the estimating means by weighting according to the influence coefficient calculated by the influence coefficient calculating means.   The emission intensity correcting unit corrects the emission intensity of each particle so that the emission amount obtained from the corrected particle emission intensity matches the emission amount estimated by the estimation unit. An apparatus for estimating the amount of released substances described. A method for estimating a release amount of a substance, which estimates a release amount of a substance released from a specific release point based on observed concentration values observed at a plurality of observation points,
Obtaining the position information of each observation point, the concentration observation value observed at the observation point and the information of the observation time, and the information acquisition process of acquiring the position information of the release point and the release start time of the substance,
An influence coefficient calculation process for calculating an influence coefficient determined according to the relative position between the position of the particle and the release point, and the relative time between the substance release time and the observation time at the observation point, using a diffusion model;
Release of particles emitted from the emission point by minimizing the residual norm expressed using the difference between the concentration observation value at the plurality of observation points and the calculated concentration value at each observation point obtained by calculation. An estimation process for estimating intensity and estimating the amount of emission from the emission point using the estimated emission intensity of the particles;
A method for estimating a release amount of a substance, comprising: a release intensity correction process that corrects the release intensity of each particle estimated in the estimation process by weighting according to the influence coefficient calculated in the influence coefficient calculation process. A substance release amount estimation program for estimating the amount of substance released from a specific release point based on concentration observation values observed at a plurality of observation points,
Obtaining the position information of each observation point, the concentration observation value observed at the observation point and the information of the observation time, information acquisition processing for acquiring the position information of the release point and the release start time of the substance,
An influence coefficient calculation process for calculating an influence coefficient determined according to the relative position between the position of the particle and the release point, and the relative time between the substance release time and the observation time at the observation point, using a diffusion model;
Release of particles emitted from the emission point by minimizing the residual norm expressed using the difference between the concentration observation value at the plurality of observation points and the calculated concentration value at each observation point obtained by calculation. An estimation process for estimating an intensity, and estimating an emission amount from the emission point using the estimated emission intensity of the particles;
A substance release amount estimation program for causing a computer to execute a release intensity correction process for correcting the emission intensity of each particle estimated in the estimation process by weighting according to the influence coefficient calculated in the influence coefficient calculation process .

Images