Measurement system and flow warning method
By arranging measurement equipment and evaluation units in the water basin and using machine learning algorithms to predict water levels and flow rates, the accuracy and advance warning problems of high water level prediction in the prior art are solved, and high-precision prediction and warning are achieved, which is suitable for Various water basins.
Patent Information
- Application Number
- CN202380036718.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-05-04
- Filing Date
- 2023-04-28
- Publication Date
- 2024-11-26
AI Technical Summary
In the latest technology, it is difficult to accurately predict the high water level situation in water basins in extreme weather events, especially when the water level and flow rate change drastically, it is impossible to issue fine warnings in advance, resulting in the inability to effectively detect high water level warning situations with limited locations.
A measurement system was designed, including the arrangement of measurement equipment in the target location of the water body and the watershed, combined with evaluation units, using machine learning algorithms (such as artificial neural networks or deep learning) to predict water levels and flow rates based on continuous measurements, and through wireless interfaces Connect with decentralized servers to enable high-precision warning prediction and transmission, and measurement equipment can be self-sufficient to reduce infrastructure and installation efforts.
Reliable prediction and advance warning of high water levels in water basins can provide sufficient pre-time in extreme weather events, detect high water levels with limited locations, and improve the prediction by combining optimization algorithms and meteorological data. Accuracy and resolution.
Smart Images

Figure CN119032300A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a measuring system and a corresponding method, with the help of which the flow rate and / or the water level in the catchment area of a water body (commonly known as the "watershed"), which is a central factor in the regional response, can be predicted so that corresponding warnings can be issued in the event of high water levels in the area. Background Art
[0002] With the increasing intensity of extreme weather events, the warning of high water level events in the basins of water bodies assumes an ever more central role. Pure recording of high water level events is based according to the prior art on level measurements at meaningful locations of the water body, such as for example:
[0003] - At the source of a water body,
[0004] - at the entrance to a body of water,
[0005] - below the level of significant inflow, with a significant increase in the area of the catchment,
[0006] - In water retention areas, natural and artificial lakes or seas with a dynamic range of water levels and their inlets and outlets.
[0007] In this case, the level can be read visually with a structure having a corresponding scale. In the case of more modern designs, the level can also be measured in an automatic manner, for example, by means of a radar-based level measurement. For example, a radar-based measuring device is known from DE 10 20 20 10 0 867 A1, by means of which a level, for example a filling level, can be determined.
[0008] A possible local analytical analysis of extreme precipitation events is described in DE 10 20 2011 9 488 A1. The measuring system is based on local measurements of precipitation at a number of measuring points and subsequent eccentricity evaluation based on a self-learning AI (“artificial intelligence”)-based algorithm. Advantageously, very small grid resolutions can be achieved.
[0009] Official warnings of high water levels are mainly based on the provision of meteorological data with the help of which the predicted precipitation is used as the central decision criterion. Based on this, reliable forecasts of the regional response in the future time period or future time point can be given within the range of lead time of several hours to several days. In this case, the main physical target variables used as regional response as a function of time are water level and flow at defined target locations (i.e. locations of interest) of the region.
[0010] The smallest possible grid with the help of which the regional response forecast can be resolved in position (usually several hundred meters) based on the pure meteorological data volume. In this case, the precipitation can vary drastically depending on the location in the case of extreme weather events. In addition, the high water level situation that may arise from this in a given case depends on the unknown water body situation in the forecast area, because the precipitation is carried away via the water body and its basin. In this case, the regional reference result of the precipitation event is called the regional response, where the water level and flow rate at the considered target location of the grid are the central, physical, measured variables of the regional response. If the - predicted - water level and / or flow rate exceeds the limit value defined individually for the target location, this is defined as a high water level situation. In this regard, not only the first-order water bodies, but also the second-order and third-order water bodies and their basins are particularly dangerous, because the precipitation is mainly carried away along these basins. Summary of the invention
[0011] The object of the present invention is to be able to provide a reliable high water level warning for at least one defined target location in the catchment area of a body of water.
[0012] The present invention achieves this object by a measuring system for predicting the water level and / or flow rate at at least one defined target location in the basin of a body of water. In this context, the term "prediction" within the scope of the present invention means calculating an expected value for a defined point in time, length of time or period of time in the future. The measuring system of the present invention comprises the following components:
[0013] - one or more measuring devices arranged at target locations in a water body and / or within a watershed in order to determine the following values in the form of measured values
[0014] oWater level
[0015] o soil moisture, such as described, for example, in WO 2019096766 A1, and / or
[0016] o precipitation, for example, the amount of precipitation accumulated or the intensity of precipitation,
[0017] - an evaluation unit connected to at least one measuring device, which is designed
[0018] oBased on at least one measurement, and
[0019] o based on a defined algorithm which describes the dependency of the flow rate and / or the water level at a target location of the body of water on at least one measured value,
[0020] To predict the flow rate and / or water level at a target location for a defined length of time or within a defined period of time.
[0021] With the measuring system of the invention, in case of suspicion, the situation can be analyzed and, in case of danger of high water levels, a warning can be issued with sufficient lead time. Furthermore, the warning can have a very fine resolution in terms of location. In this way, warning situations which are greatly restricted in terms of location can also be detected, which would not be detectable with the aid of conventional forecasts based on purely meteorological data. In particular, when the water level / flow rate and, in a given case, the direction of flow at the respective target location are predicted by a measuring system covering a defined time period, the prediction result for a specific point in time can also be correspondingly corrected, i.e. updated, in a given case.
[0022] The measuring system of the invention can be designed particularly effectively when a machine learning algorithm, in particular a machine learning algorithm formed as an artificial neural network or a deep learning method, is implemented in the evaluation unit. In this case, the algorithm can be trained or optimized, in particular based on continuously measured measured values and based on a water level measured at least once at the target location and / or a flow rate measured at least once at the target location.
[0023] In order that the measuring system according to the invention requires as little infrastructure as possible and can be installed with as little effort as possible, it is advantageous if one or more measuring devices are designed to be energy self-sufficient, for example by equipping them with corresponding buffer batteries. In particular, in order to keep the battery life as long as possible, the measuring rate of the measuring device or its transmission rate to the evaluation unit can be controlled proportionally or stepwise as a function of the measured value or the predicted flow rate or the predicted water level. In this way, in the event of a high water level event that may be reached, precisely in the event of exceeding a defined limit value, it is ensured that the measuring device measures and transmits the measured value frequently enough. When the measured value of the measuring device or the corresponding predicted area reaction (water level / flow rate) does not exceed the limit value, for example, the measurement and transmission rate can be correspondingly slowed down by the evaluation unit control.
[0024] In general, the evaluation unit of the measuring system of the invention can be implemented by means of any suitable digital circuit, such as, for example, an FPGA, a microcontroller or a storage medium cooperating with a corresponding program. Preferably, a decentralized server can be used as the evaluation unit, wherein at least one measuring device is connected to the server via a wireless interface, in particular a GSM interface. This ensures a fast and secure transmission of the measured values.
[0025] Alternatively, in addition, meteorological forecast data of the basin of the water body can be combined to optimize the predicted water level and / or the predicted flow rate at the target location. For this reason, it is necessary to design the evaluation unit accordingly so that such data can be received. This can again occur, for example, via a GSM interface. In this case, the algorithm must be designed accordingly so that the meteorological forecast data is combined into the calculation of the predicted flow rate and / or the predicted water level at the target location.
[0026] Corresponding to the measuring system of the invention, the object of the invention is achieved by a matching measuring method. In this case, the method is used to predict the flow rate and / or water level at at least one defined target location in the flow area of a water body by means of a measuring system according to one of the above-described embodiments. The method comprises the following method steps:
[0027] - Record as at least one measurement at the target location in the water body and / or within the catchment area at least the following
[0028] oWater level
[0029] oSoil moisture, and / or
[0030] oPrecipitation, and
[0031] - calculating at least one target location a flow rate predicted for a defined length of time and / or a water level predicted for that length of time based on at least
[0032] oat least one measurement, and
[0033] oAlgorithms.
[0034] In this case, the high water warning of the invention can be improved by determining the water level and / or the flow rate not only for one target location but for two or more different target locations in the basin. In this case, a "virtual" arrangement of the target locations by the algorithm can take place in a grid form. Especially in the case of a server-based design of the evaluation unit, the correspondingly increased computational effort to realize a plurality of target locations is not a problem. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] The invention will now be explained in more detail based on the accompanying drawings, the only figure of which is illustrated as follows:
[0036] Figure 1 A schematic diagram of a watershed that is part of a flowing body of water. DETAILED DESCRIPTION
[0037] In order to understand the high water level prediction principle of the present invention, Figure 1 Schematically shown is a portion of a flowing water body 2, including its watershed 1. In this case, the term "watershed" refers to an area where hydrological factors influence the water level and / or flow of the flowing water body 2. Thus, factors counted are, for example, the tributaries of the water body in area 1, the topography and the water-holding capacity of the soil. "Water bodies" include not only flowing water bodies such as streams and rivers, but also, for example, temporarily dry riverbeds and stationary water bodies such as lakes and seas.
[0038] For high water level monitoring in the basin 1, this can within the scope of the invention be divided into grids with a grid size of, for example, between 10x10m and 100x100m, such as in Figure 1 In this case, each of the grid points forms a potential target location x, which can be considered for describing the regional response. This means that for each target location x, a possible water level L, a possible flow rate and the direction of emission in a given case as a function of various input variables.
[0039] exist Figure 1 In the case of the view shown, for reasons of clarity, only one of the grid points is defined as the target position x. Contrary to this simplified view, it is possible within the scope of the invention that the target positions x of at least some of the grid arrangements will be located outside the flowing water body 2. In these cases, the water level L and the flow rate are therefore In the absence of high water conditions, the water level L and / or flow rate are predicted to be 0 L and 0 L / sec, respectively. Above a limit value L defined individually for such a target position max , , this means there is a high water warning condition.
[0040] In order to predict regional reactions and thus possible high water situations as a reaction to precipitation events in the basin 1, at least the following aspects are recorded as measured values according to the invention at appropriate measuring points:
[0041] oInstantaneous water level
[0042] oCurrent soil moisture, and / or
[0043] oPrecipitation, such as the amount of precipitation that has fallen to date, or the intensity of precipitation
[0044] exist Figure 1 In the case of the schematic diagram, the measuring system of the invention comprises a measuring device type for recording corresponding measured values in the watershed 1, thus,
[0045] - a water level measuring device 10 located in a branch of the flowing water body 2,
[0046] - a water level measuring device 10 ′ in the main channel of the flowing water body 2 upstream of the target position x,
[0047] - a moisture measuring device 11 for determining soil moisture, and
[0048] - A precipitation measuring device 12 which can record at least instantaneous precipitation as a binary event (yes / no).
[0049] In this case, it is clear that the regional response - hence the grid density of the target locations x and the water level L they identify, as well as the flow rate , it can be predicted in more detail and safely that more measuring devices 10, 10', 11, 12 and measuring device types are applied by the measuring system in the basin 1. In order that the measuring devices 10, 10', 11, 12 can work automatically and wirelessly, it is advantageous to equip them with corresponding buffer batteries and / or additional solar cells in each case. This also facilitates the relocation of the measuring devices 10, 10', 11, 12 in the basin 1.
[0050] The measured values recorded by the measuring devices 10, 10', 11, 12 are fed to the evaluation unit 3 of the measuring system as input variables of the corresponding algorithm. In this case, the evaluation unit 3 corresponds to Figure 1 The embodiment shown is preferably implemented decentralized and is not part of one of the measuring devices 10, 10', 11, 12. In this way, in the event of a possible failure of one of the measuring devices 10, 10', 11, 12, the high water level warning is not lost. Figure 1 In the case of the embodiment shown, the evaluation unit 3 is implemented in the form of a decentralized server which is addressable, for example, via a GSM interface. Correspondingly, the measuring devices 10, 10', 11, 12 of the measuring system of the invention also have corresponding wireless interfaces in order to be able to communicate with the server-based evaluation unit 3. In this connection, the transmission rate of the measuring devices 10, 10', 11, 12 to the evaluation unit 3 can be controlled by these according to the currently measured measured values in each case or according to the predicted regional reaction. When the current measured values and / or the predicted flow rate The transmission rate or the measuring rate can also be increased when the water level L is predicted to increase, for example, above a defined limit value. In this case, the increase can take place stepwise or almost linearly. Advantageously, the measuring rate and / or the transmission rate of the individual measuring devices 10, 10', 11, 12 can be set to a very low default value when no warning situation is expected. In this way, the readiness of the measuring devices 10, 10', 11, 12 operating autonomously can be significantly prolonged.
[0051] The algorithm implemented in the evaluation unit 3 describes the following relationship.
[0052] o the measured values obtained from the measuring devices 10, 10', 11, 12, and
[0053] o Predicted flow rate at target location x in flow basin 1 and / or water level L.
[0054] In this case, it is clear that, depending on the design of the algorithm, not every target position x has to be arranged on the grid. Instead, one or more target positions x can also be distributed individually according to the specific watershed 1. Likewise, in the context of the present invention, depending on the situation in the specific watershed 1, it is also possible to choose to arrange at least some of the measuring devices (types) 10, 10', 11, 12 on the grid or at the determined target positions x.
[0055] In order to make the measurement system of the present invention applicable to various water bodies 2 and various river basins 1, the algorithm must be combined
[0056] - the topology of watershed 1,
[0057] - its soil characteristics, and
[0058] - The extent of flowing water bodies2 and their tributaries and discharges.
[0059] An opportunity to take these factors into account is to design the algorithm in the evaluation unit 3 as a machine learning algorithm. For this purpose, a corresponding artificial neural network can be formed in the evaluation unit 3, for example in the form of "deep learning". The target variables to be optimized here include in particular the accuracy and safety of the prediction of the regional reaction, and therefore the water level L and the flow rate at the respective target location x. The prediction accuracy and prediction safety of , as well as the best resolution or position corresponding to the target position x in a given situation.
[0060] For this optimization, on the one hand, the measured values ascertained by the measuring devices 10, 10', 11, 12 must be transmitted to the evaluation unit 3 and thus to the algorithm, and in fact, as continuously as possible during at least a defined initial learning phase. On the other hand, at least during this learning phase, the corresponding water level L and flow rate must be transmitted to the evaluation unit 3 and thus to the algorithm. is supplied to the evaluation unit 3 and thus to the machine learning algorithm. In this case, within the scope of the invention itself, the values L, Whether it comes from a measurement at the corresponding target position x, or whether a suitable text or estimate is given, is not itself decisive. Furthermore, whether the machine learning algorithm is designed to use "supervised" learning or "unsupervised" learning is not decisive. Furthermore, the initial learning phase does not need to be time-limited, so that the algorithm can instead be continuously optimized or adapted during the overall operation of the measurement system.
[0061] and Figure 1Compared to the embodiment shown in , the measuring system of the invention can be applied not only to flowing water bodies 2, but also alternatively to all other water bodies whose basins are at risk of high water levels. Overall, the high water level prediction of the invention with the aid of the measuring system offers the advantage that, on the one hand, very timely predictions of flow rates and water levels can be given. Thus, in the case of suspicion, the situation can be analyzed and, in the case of a high water level risk, a warning can be issued with sufficient lead time. On the other hand, a warning can be given with a very large local resolution. In this way, possible warning situations that are very restricted in location can also be detected, which would not be detected with the aid of conventional predictions based on pure meteorological forecast data. However, it is also an option within the scope of the invention to incorporate such meteorological forecast data into the predicted water level / flow rate of the invention, when these data are correspondingly transmitted to the evaluation unit 3, or when the implemented algorithm correspondingly incorporates such forecast data. In the case of a self-learning algorithm, this can be used in a supplementary manner to better fit the high water level to a specific basin (1).
[0062] Reference numerals list
[0063] 1 Basin
[0064] 2 Water bodies
[0065] 3 Evaluation Units
[0066] 10, 10' water level measuring equipment
[0067] 11Soil moisture measurement equipment
[0068] 12Precipitation measurement equipment
[0069] L water level
[0070] Flow rate
[0071] X target position
Claims
1. A method for predicting a water level (L) and / or a flow rate (L) at at least one defined target location (x) in a watershed (2) of a body of water (3). ) measurement system, including: - at least one measuring device (10, 10', 11, 12) arranged at the target location (x) of the body of water (2) and / or in the watershed (1) in order to determine at least the following as measured values oWater level oSoil moisture, and / or oPrecipitation, - an evaluation unit (3) connected to the at least one measuring device (10, 10', 11, 12), the evaluation unit being designed as oBased on at least one measurement, and o Based on a defined algorithm, the defined algorithm describes the flow rate ( ) and / or the dependence of said water level (L) on said at least one measured value, to predict the flow rate ( ) and / or the water level (L).
2. The measurement system according to claim 1, wherein: A machine learning algorithm is formed in the evaluation unit (3) so that the algorithm can be trained based on the following items: o measured values continuously measured by the measuring device (10, 10', 11, 12), and o a water level (L) measured at least once at the target location (x) and / or a flow rate ( ).
3. The measuring system according to claim 2, wherein: The machine learning algorithm is designed as an artificial neural network, in particular in the form of a deep learning method.
4. The measuring system according to claim 1, wherein: The evaluation unit (3) is designed to predict the flow rate ( ) and / or the water level (L).
5. The measuring system according to claim 1, wherein: The at least one measuring device (10, 10', 11, 12) is designed in particular to set a measuring rate or a transmission rate to the evaluation unit (3) proportionally or stepwise according to oMeasured value, or oPredicted flow rate ( ) or predicted water level (L).
6. The measuring system according to claim 1, wherein: The evaluation unit (3) is designed as a decentralized server, and the at least one measuring device (10, 10', 11, 12) is connected to the server (3) via a wireless interface, in particular a GSM interface.
7. The measurement system according to claim 6, wherein: The at least one measuring device (10, 10', 11, 12) is designed to be energy-autonomous, in particular battery-operated.
8. The measuring system according to at least one of the preceding claims, wherein The evaluation unit (3) is designed to receive meteorological forecast data for the basin (2) of the water body (2), and wherein the algorithm is designed to incorporate the meteorological forecast data into the predicted flow rate ( ) and / or the calculation of the predicted water level (L).
9. A method for predicting a flow rate (x) at at least one defined target location (x) in a flow basin (1) of a body of water (2) by means of a measuring system according to one of the preceding claims. ) and / or water level (L), where: The method comprises the following steps: - recording as at least one measurement value at the target location (x) of the water body (2) and / or within the watershed (1) at least the following oWater level oSoil moisture, and / or oPrecipitation, and - Calculate a predicted flow rate for a defined length of time at at least one target location (x) based on at least ) and / or the water level predicted for that length of time (L) o the at least one measured value, and oThe algorithm described.
10. The method according to claim 9, wherein: The water level (L) and / or the flow rate (L) are determined at at least four different target locations (x) in the flow basin (1), in particular at locations arranged in a defined grid relative to one another. ).
Citation Information
Patent Citations
Mounting device for level measuring devices
DE102020100867A1
Method and system for analyzing precipitation events, as well as computer program product and use
DE102020119488A1
Device for determining the moisture and / or the conductivity of a medium
WO2019096766A1
