阅读说明：本技术 基于边缘计算的闸前反射式水位计智能校准方法及系统 (Intelligent calibration method and system for front-gate reflection type water level meter based on edge calculation ) 是由 方卫华 孙勇 梅星 蔡悦 卢涛 蒋涛 张威 于 2021-07-22 设计创作，主要内容包括：本发明公开一种基于边缘计算的闸前反射式水位计智能校准方法及系统,采用现场嵌入式计算机、水位高精度率定平台、多要素传感器结合热力耦合理论和机器学习方法,采用定期离线建立测值误差修正模型和在线实时误差修正策略,实现了闸前水位的智能校准。离线校准首先固定水位基准面,然后控制水位测量装置及支架在高精密丝杆上以预定的高度间隔上下移动,同时测量水位值、气象参数和水面至支架之间透明度,利用样本构建测值误差修正模型。在线测量时将影响因素作为测值误差修正模型的输入进行实时误差修正。本方法直观严密、系统先进可靠,可推广到其他对水位测量精度和长期稳定性要求高的应用场景。(The invention discloses an intelligent calibration method and system for a reflection type water level meter before a gate based on edge calculation. The off-line calibration firstly fixes a water level datum plane, then controls the water level measuring device and the bracket to move up and down on the high-precision screw rod at a preset height interval, simultaneously measures the water level value, meteorological parameters and transparency between the water surface and the bracket, and utilizes a sample to construct a measured value error correction model. And during online measurement, the influence factors are used as the input of the measured value error correction model to carry out real-time error correction. The method is visual and strict, the system is advanced and reliable, and the method can be popularized to other application scenes with high requirements on water level measurement accuracy and long-term stability.)

1. The intelligent calibration method of the reflection type water level gauge before the gate based on the edge calculation is characterized by comprising an off-line calibration stage and further comprising the following steps:

the lifting control motor controls the water level tray to be horizontally placed, the full water level of the lifted water level tray is higher than the water level before the gate and is fixed, and a reference surface is provided for the calibration of the water level before the gate;

the stepping motor controls the screw rod to rotate, drives the water level measuring device and the support which are connected with the screw rod to move upwards at a preset interval height from an initial position, measures the water level and meteorological parameters before the water level measuring device and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move a preset interval, and records the data until the water level measuring device and the support move upwards to the top end;

the stepping motor controls the screw rod to rotate reversely, drives the water level measuring device and the support which are connected with the screw rod to move downwards at a preset interval height, measures the water level and meteorological parameters before the water gate and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move a preset interval, and records the data until the water level measuring device and the support move downwards to the lowest end;

the stepping motor controls the screw rod to rotate reversely again, drives the water level measuring device and the support which are connected with the screw rod to move upwards at preset intervals, measures the water level and meteorological parameters before the water level measuring device and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move for a preset interval, records the data until the water level measuring device and the support move upwards to the middle of the measuring range of the reflection type water level meter, and stops moving the water level measuring device and the support;

establishing a plurality of measured value error correction models, taking a plurality of groups of measured water level before gate, meteorological parameters and transparency data as input samples, taking the difference between a water level value corresponding to the moving distance of a water level measuring device and a bracket on a screw rod and a water level value before gate as an output sample, respectively adding the input samples and the output samples into the plurality of measured value error correction models, training the plurality of measured value error correction models to obtain a plurality of trained measured value error correction models, comprehensively evaluating the trained measured value error correction models, and determining an optimal measured value error correction model;

and carrying out real-time error correction on the measured value of the reflective water level meter by using the determined optimal measured value error correction model, and carrying out online measurement on the water level before the gate.

2. The method for intelligent calibration of a front-gate reflection type water level gauge based on edge calculation according to claim 1, wherein the step of online measurement of the front-gate water level comprises: inputting the pre-gate water level value and the time synchronization meteorological and transparency data into a selected optimal measurement value error correction model, correcting errors through edge calculation, outputting a calibrated pre-gate water level value on line, and locally storing and/or transmitting the calibrated pre-gate water level value to a rear processing center or an authorized mobile terminal.

3. The method of claim 1, wherein the video image is combined with weather and operation conditions of the sluice to determine whether the water level before the sluice is stable, and if the water level before the sluice is stable, the water level before the sluice is directly used as a reference surface for calibration.

4. The intelligent calibration method for the edge-calculation-based front-gate reflective water level gauge according to claim 1, wherein an underwater tray is used as a reference surface for calibration, and after the determination of the optimal measurement value error correction model is completed, the underwater tray is erected and retracted until the underwater tray is tightly attached to the wall surface where the upper vertical guide rail and the lower vertical guide rail are located, and then the front-gate water level is measured on line.

5. The intelligent calibration method for the edge-computing-based front-gate reflection type water level gauge of claim 1, wherein during the calibration process, the water level measuring device and the bracket move up and down periodically within the measuring range of the reflection type water level gauge; and when the calibration is finished, the staying and fixing positions of the water level measuring device and the bracket are the middle part of the measuring range of the reflection type water level meter.

6. The method of claim 1, wherein a correlation vector machine, a long-time memory, a coding-decoding neural network model and a gate control cycle unit are selected as the measurement error correction model.

7. The method for intelligent calibration of a front gate reflection type water level gauge based on edge calculation as claimed in claim 1, wherein the method for comprehensive evaluation of the trained measured value error correction model comprises: calculating the measurement of errors in a predicted value obtained by measuring the value error correction model by randomly extracting a plurality of groups of input combinations, and judging the precision of the value error correction model; adding a certain error into an input sample, calculating the error in a predicted value obtained by a measured value error correction model, and judging the stability of the measured value error correction model; judging the generalization capability of the measured value error correction model by judging whether the corresponding time length meets the requirement when the prediction accuracy is the same maximum allowable error; and comprehensively considering the model with the best precision, stability and generalization capability as the determined optimal measured value error correction model.

8. A system for intelligent calibration of a reflection type water level gauge in front of a gate based on edge calculation is used for realizing the method of any one of claims 1 to 7, and is characterized by comprising a measuring device and a bracket, a bracket lifting control device, a datum plane providing device, a measuring communication control device, a leveling punctuation and a reinforced concrete pier surface, wherein the measuring device and the bracket are used for measuring the water level in front of the gate and meteorological and transparency data; the support lifting control device is used for driving the water level measuring device and the support to move up and down at a fixed interval height and to be fixed; the reference surface providing device is used for providing a stable water surface reference during calibration and ensuring that the water flow and the water surface are not influenced during measurement; the measurement communication control device is connected with the measurement device, the support lifting control device and the reference surface providing device and is used for measuring the parameters of the sensor, the control motor, the edge calculation, the front-back communication, the storage and the recording of relevant measured data, programs and settings.

9. The method for intelligent calibration of a front gate reflection type water level gauge based on edge calculation of claim 8, wherein the support lifting control device comprises a driving motor, and the driving motor drives the screw rod to rotate in forward and reverse directions.

Technical Field

The invention relates to the field of hydraulic engineering, in particular to an intelligent calibration method and system for a front-gate reflection type water level meter based on edge calculation.

Background

The water level before the water gate is an important parameter for water gate operation control, structure safety analysis and flow calculation, and the water gate has low water head, rapid water level change and sensitive calculation and analysis model to the water level, so the water gate has high requirement on water level observation precision, the existing reflection type water level gauge is easily influenced by environmental temperature, atmospheric humidity, water surface transpiration, flow water mist and the like to cause measurement errors, and the temperature drift and the time drift of the water level gauge are added, so the error is larger, and the trend of gradual development is increased along with the time.

With the development of science and technology, intelligent algorithms and models, particularly the rapid development of embedded systems, make it possible to adopt embedded edge calculation for the detected water level, and the method is used for solving the field calibration problem of the reflection type water level gauge in front of the gate in the prior art.

Through retrieval, China with the publication number of CN108020272A specially facilitates 2018, 5.11.A device for online monitoring of open channel gate crossing flow is disclosed, which comprises an online monitor, a water level gauge in front of a gate, a water level gauge behind the gate, a gate opening gauge, a box door state sensor, a battery, a mobile terminal and a server, wherein the online monitor is respectively connected with the water level gauge in front of the gate, the water level gauge behind the gate, the gate opening gauge, the box door state sensor and the battery through a wired communication circuit; the online monitor includes a low power consumption power module. The invention realizes high integration, low power consumption, safety and theft prevention of the open channel gate flow metering device and provides comprehensive and accurate metering data for realizing water-saving irrigation. However, the invention does not calibrate the data detected by the water gate water level gauge, and the technical problem proposed by the application still exists.

Through retrieval and discovery, China with the publication number of CN205808494U specially used for 2016, 12 and 14 discloses a comprehensive monitoring system for water level before gate, which comprises a water level sensor, an analog-digital converter, a microprocessor, a data transmission module, a man-machine exchange server, a gate device, a solar cell module and a storage battery device; the water level sensor collects water level data in front of the gate and transmits the water level data to the microprocessor through the analog-to-digital converter, the data transmission module can realize two-way communication with the human-computer interaction server, the gate device realizes remote control of opening and closing of the gate through receiving instructions of the human-computer interaction server, and the solar cell module is connected with the storage battery device and supplies power for all devices except the human-computer interaction server. The utility model utilizes the water level sensor to obtain the real-time water level before the floodgate, and is combined with the flood peak forecast, thereby realizing the comprehensive monitoring of the water level before the floodgate; the intelligent flood control system has the advantages of small space occupation and high intelligent degree, can monitor and control the water level in front of the gate for a long time, achieves timely, accurate and safe flood control, and ensures the safety of people's life and materials. However, the utility model discloses a still do not consider the error of environmental impact to the water level data before the floodgate that level sensor gathered, still exist the technical problem that this application provided.

Therefore, the reduction of the error of the water level data before the gate detected by the reflection type water level meter is still a technical problem which is urgently solved by the technical personnel in the field.

Disclosure of Invention

In order to overcome the defects of the prior art, the invention provides an intelligent calibration method and system of a front-gate reflection type water level gauge based on edge calculation, and solves the problem of field calibration of the front-gate reflection type water level gauge.

The invention is realized by the following technical scheme:

the intelligent calibration method of the reflection type water level gauge before the gate based on the edge calculation comprises an off-line calibration stage and further comprises the following steps:

the lifting control motor controls the water level tray to be horizontally placed, and the full water level of the water level tray is lifted to be higher than the water level before the gate and is fixed, so that a reference surface is provided for the calibration of the water level before the gate;

the stepping motor controls the screw rod to rotate, drives the water level measuring device and the support which are connected with the screw rod to move upwards at a preset interval height from an initial position, measures the water level and meteorological parameters before the water level measuring device and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move a preset interval, and records the data until the water level measuring device and the support move upwards to the top end;

the stepping motor controls the screw rod to rotate reversely, drives the water level measuring device and the support which are connected with the screw rod to move downwards at a preset interval height, measures the water level and meteorological parameters before the water gate and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move a preset interval, and records the data until the water level measuring device and the support move downwards to the lowest end;

the stepping motor controls the screw rod to rotate reversely again, drives the water level measuring device and the support which are connected with the screw rod to move upwards at preset intervals, measures the water level and meteorological parameters before the water level measuring device and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move for a preset interval, records the data until the water level measuring device and the support move upwards to the middle of the measuring range of the reflection type water level meter, and stops moving the water level measuring device and the support;

establishing a plurality of measured value error correction models, taking a plurality of groups of measured water level before gate, meteorological parameters and transparency data as input samples, taking the difference between a water level value corresponding to the moving distance of a water level measuring device and a bracket on a screw rod and a water level value before gate as an output sample, respectively adding the input samples and the output samples into the plurality of measured value error correction models, training the plurality of measured value error correction models to obtain a plurality of trained measured value error correction models, comprehensively evaluating the trained measured value error correction models, and determining an optimal measured value error correction model;

and carrying out real-time error correction on the measured value of the reflective water level meter by using the determined optimal measured value error correction model, and carrying out online measurement on the water level before the gate.

In the technical scheme, the off-line calibration firstly fixes a water level datum plane, after the datum plane is determined, the water level measuring device and the support are controlled to move up and down on the screw rod at a preset height interval, the water level value, the meteorological parameters and the air transparency between the water level and the water level measuring device and between the water level and the support are measured at the same time, a measured value error correction model is constructed by using samples, the measured value error correction model is trained by using input samples and output sample data to obtain a trained measured value error correction model, the trained measured value error correction model is comprehensively evaluated, and the optimal measured value error correction model is determined. The method is visual and strict, the system is advanced and reliable, and the method can be popularized to other application scenes with high requirements on water level measurement accuracy and long-term stability.

Specifically, the full water level of the underwater tray which is lifted to a level higher than the water level before the gate by the lifting control motor provides a reference surface which is stable during each calibration process and is a constant water level.

Specifically, the meteorological parameters further include wind speed, humidity, air pressure, and air temperature data.

Specifically, the screw rod is a high-precision screw rod, the transmission high-precision screw rod is provided with a stainless steel trapezoidal screw rod, and a pressure spring is arranged in a matched spiral pair to eliminate clearance errors in movement.

Preferably, the step of online measuring the water level before the gate comprises: inputting the pre-gate water level value and the time synchronization meteorological and transparency data into a selected optimal measurement value error correction model, correcting errors through edge calculation, outputting a calibrated pre-gate water level value on line, and locally storing and/or transmitting the calibrated pre-gate water level value to a rear processing center or an authorized mobile terminal.

Preferably, the water level before the sluice is judged to be stable or not according to the video image and the weather of the upstream and downstream of the sluice and the running condition of the sluice, and if the water level before the sluice is judged to be stable, the water level before the sluice is directly used as a reference surface for calibration.

Preferably, the underwater tray is used as a reference surface for calibration, after the optimal measurement value error correction model is determined, the underwater tray is erected and folded until the underwater tray is tightly attached to the wall surface where the upper vertical guide rail and the lower vertical guide rail are located, and then the water level before the gate is measured on line.

Preferably, during the calibration process, the water level measuring device and the bracket move up and down periodically within the measuring range of the reflection type water level meter; and when the calibration is finished, the staying and fixing positions of the water level measuring device and the bracket are the middle part of the measuring range of the reflection type water level meter.

Preferably, a correlation vector machine, a long-time memory, an encoding-decoding neural network model and a gate control cycle unit are selected as the measurement value error correction model.

Preferably, the method for comprehensively evaluating the trained measured value error correction model includes: calculating the measurement of errors in a predicted value obtained by measuring the value error correction model by randomly extracting a plurality of groups of input combinations, and judging the precision of the value error correction model; adding a certain error into an input sample, calculating the error in a predicted value obtained by a measured value error correction model, and judging the stability of the measured value error correction model; judging the generalization capability of the measured value error correction model by judging whether the corresponding time length meets the requirement when the prediction accuracy is the same maximum allowable error; and comprehensively considering the model with the best precision, stability and generalization capability as the determined optimal measured value error correction model.

Specifically, the length of the motor-driven high-precision screw rod (considering temperature correction) is determined as a true value, the measured value of the reflection type water level measurement is a measured value containing errors, and the measured value has certain errors due to drift caused by instrument stability of the reflection type water level meter, influence of a measuring medium and instability of a measuring system. The driving motor rotates the control bracket to move up and down according to fixed steps, temperature correction is considered according to the thread pitch of the high-precision screw rod in each step and then the temperature correction is multiplied by the steps, and the real value H of each step can be obtained by combining the water surface elevation of the tray_i(i 1, 2, 3.., N), and the measured value acquired by the reflection type water level meter at the same time is h_i(i ═ 1, 2, 3.., N), there is a relation H_i＝h_i+v_iIf it is true, a "correction value" for the "measured value" required for each step is obtained as v_i＝H_i-h_iObtained by a cycle"correction value" sequences v_i(i ═ 1, 2, 3.., N). By analysis of v_iThe influence of (i ═ 1, 2, 3.., N) can be easily found, and the following relationship v exists_i＝f(T_i，t，h_i，M_i，τ_i，ω_i，w_i) Wherein, T_i、t、h_i、M_i、τ_i、ω_i、w_iRespectively representing a temperature measurement, a time, a reflection level measurement, a humidity measurement, a transparency of air from the water surface to the reflection level, a wind force and a wind direction. The measured values are obtained by synchronous measurement of a system local sensor. Since f (-) is a nonlinear function, nonlinear neural network model approximation, such as a correlation vector machine (SVM), long-time memory (LSTM), encoding-decoding neural network model and gated round robin unit (GRU), is used as the measured value error correction model. Obtaining stable, reliable and high-precision v after periodically calibrating the model_i＝f(T_i，t，h_i，M_i，τ_i，ω_i，w_i) And the corrected water level value before the gate obtained in daily measurement is as follows:

h (t) ═ h (t) + v (t), where h (t) is the reflection type water level measurement value after correction by edge calculation, h (t) is the reflection type water level measurement value actually measured output before correction, and v (t) is the output of the measured value error correction model which is good in calibration.

The intelligent calibration system of the reflection type water level gauge in front of the gate based on the edge calculation comprises a measuring device and a support, a support lifting control device, a datum plane providing device, a measuring communication control device, a level mark point and a reinforced concrete pier surface, wherein the measuring device and the support are used for measuring water level in front of the gate and meteorological and transparency data; the support lifting control device is used for driving the water level measuring device and the support to move up and down at a fixed interval height and to be fixed; the reference surface providing device is used for providing a stable water surface reference during calibration and ensuring that the water flow and the water surface are not influenced during measurement; the measurement communication control device is connected with the measurement device, the support lifting control device and the reference surface providing device and is used for measuring the parameters of the sensor, the control motor, the edge calculation, the front-back communication, the storage and the recording of relevant measured data, programs and settings.

Preferably, the support lifting control device comprises a driving motor, and the driving motor drives the screw rod to rotate forwards and reversely.

Compared with the prior art, the invention has the beneficial effects that:

(1) during off-line calibration, firstly, a water level datum plane is fixed, after the datum plane is determined, the water level measuring device and the support are controlled to move up and down on the screw rod at a preset height interval, the water level value, meteorological parameters and the transparency of air from the water level to the water level measuring device and the support are measured at the same time, a measured value error correction model is constructed by using samples, the measured value error correction model is trained by using input samples and output sample data, the trained measured value error correction model is obtained, comprehensive evaluation is carried out on the trained measured value error correction model, the optimal measured value error correction model is determined, and the problem of field calibration of the reflection type water level meter in front of the gate is solved.

(2) The invention can periodically update the established measuring value error correction model and output the water surface elevation after the calibration on line according to the measuring value error correction model.

(3) The measured value error correction model established by the invention analyzes the measurement sensitivity, selects the position with highest precision and/or strongest anti-interference capability, and reduces the influence of the environment on the detection.

Drawings

FIG. 1 is a flow diagram of a method according to an embodiment of the invention;

FIG. 2 is a system equipment diagram according to an embodiment of the present invention;

in the figure: 1. a measuring device and a support; 2. a precise lifting control device of the bracket; 3. a reference surface providing device; 4. a measurement communication control device; 5. leveling punctuation; 6. a reinforced concrete pier surface; 1-1, a reflection type water level meter; 1-2, a video and water mist concentration probe; 1-3, a lifting nut sample driving component with internal threads; 1-4, positioning a sliding sleeve part smoothly on the hollow inner wall of the cylinder; 1-5, a small weather station; 1-6, cantilever rigid rod; 2-1, driving a motor; 2-2, a tilt sensor; 2-3, fixing and positioning a rigid support by a motor; 2-4, positioning a smooth guide rod; 2-5, high-precision lead screw; 2-6, a vertical bearing; 2-7, an adjustable base; 3-1, an underwater tray; 3-2, a positioning rotating shaft can be controlled; 3-3, anchoring the rod piece; 3-4, lifting the guide rail up and down; 3-5 lifting and controlling the motor.

Detailed Description

The technical solutions of the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings, and it is to be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without any inventive step, are within the scope of the present invention.

In the description of the present invention, it is to be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", and the like, indicate orientations and positional relationships based on those shown in the drawings, and are used only for convenience of description and simplicity of description, and do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be considered as limiting the present invention. Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.

In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; may be mechanically connected, may be electrically connected or may be in communication with each other; either directly or indirectly through intervening media, either internally or in any other relationship. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.

In the present invention, unless otherwise expressly stated or limited, "above" or "below" a first feature means that the first and second features are in direct contact, or that the first and second features are not in direct contact but are in contact with each other via another feature therebetween. Also, the first feature being "on," "above" and "over" the second feature includes the first feature being directly on and obliquely above the second feature, or merely indicating that the first feature is at a higher level than the second feature. A first feature being "under," "below," and "beneath" a second feature includes the first feature being directly under and obliquely below the second feature, or simply meaning that the first feature is at a lesser elevation than the second feature.

Example 1

As shown in fig. 1-2, the present embodiment provides an intelligent calibration method and system for a reflective water level gauge before a gate based on edge calculation, which uses a field embedded computer, a high-precision water level fixing platform, a multi-factor sensor, a thermodynamic coupling theory and a machine learning method, and simultaneously senses output signals and environmental information of the water level gauge, and adopts a strategy of periodically establishing a measured value error correction model offline and online real-time output, thereby realizing intelligent calibration of the water level before the gate.

The method can be popularized to the occasions of applying the reflective water level meter, such as behind a gate, in front of a dam, in a channel and the like, and the high-precision monitoring of the water level is realized.

Carrying out self-checking on the system equipment, judging whether the system equipment is normal or not, and if not, repairing the system equipment until the system equipment is normal; if the inclination of the vertical rod is normal, checking the inclination of the vertical rod, and if the inclination of the vertical rod does not meet the requirement, vertically adjusting until the inclination of the vertical rod meets the requirement; if the inclination of the vertical rod meets the requirement, the accuracy of the measured value error correction model is judged, and if the accuracy of the measured value error correction model does not meet the requirement, the measured value error correction model is calibrated; if the calibration of the measured value error correction model does not meet the requirement, the measured value error correction model is calibrated again until the precision of the measured value error correction model meets the requirement; and using the optimal measured value error correction model meeting the precision requirement to measure the water level.

The intelligent calibration method of the reflection type water level gauge before the gate based on the edge calculation comprises an off-line calibration stage and further comprises the following steps:

the lifting control motor controls the water level tray to be horizontally placed, and the full water level of the water level tray is lifted to be higher than the water level before the gate and is fixed, so that a reference surface is provided for the calibration of the water level before the gate; the stepping motor controls the screw rod to rotate, drives the water level measuring device and the support which are connected with the screw rod to move upwards at a preset interval height from an initial position, measures the water level and meteorological parameters before the water level measuring device and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move a preset interval, and records the data until the water level measuring device and the support move upwards to the top end; the stepping motor controls the screw rod to rotate reversely, drives the water level measuring device and the support which are connected with the screw rod to move downwards at a preset interval height, measures the water level and meteorological parameters before the water gate and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move a preset interval, and records the data until the water level measuring device and the support move downwards to the lowest end; the stepping motor controls the screw rod to rotate reversely again, drives the water level measuring device and the support which are connected with the screw rod to move upwards at preset intervals, measures the water level and meteorological parameters before the water level measuring device and the air transparency data from the water level measuring device and the support to the water surface once every time the water level measuring device and the support move for a preset interval, records the data until the water level measuring device and the support move upwards to the middle of the measuring range of the reflection type water level meter, and stops moving the water level measuring device and the support; establishing a plurality of measured value error correction models, taking a plurality of groups of measured water level before gate, meteorological parameters and transparency data as input samples, taking the difference between a water level value corresponding to the moving distance of a water level measuring device and a bracket on a screw rod and a water level value before gate as an output sample, respectively adding the input samples and the output samples into the plurality of measured value error correction models, training the plurality of measured value error correction models to obtain a plurality of trained measured value error correction models, comprehensively evaluating the trained measured value error correction models, and determining an optimal measured value error correction model; and carrying out real-time error correction on the measured value of the reflective water level meter by using the determined optimal measured value error correction model, and carrying out online measurement on the water level before the gate.

Further, the water level measuring device and the bracket are moved upward or downward at a fixed interval height, which is generally set to 5 cm.

Further, the meteorological parameters further include wind speed, humidity, air pressure, and air temperature data.

Furthermore, the water level measuring device and the bracket are arranged at a preset position Y, and the Y is the current preset position of the water level measuring device and the bracket.

Further, the water level measuring device and the bracket are arranged at a preset position Z, and Z is the ever or highest position of the water level measuring device and the bracket.

Further, the water surface elevation X is calibrated and kept stable during the whole calibration process.

As an embodiment, the step of online measuring the water level before the gate comprises: inputting the pre-gate water level value and the time synchronization meteorological and transparency data into a selected optimal measurement value error correction model, correcting errors through edge calculation, outputting a calibrated pre-gate water level value on line, and locally storing and/or transmitting the calibrated pre-gate water level value to a rear processing center or an authorized mobile terminal.

In one embodiment, the video image is combined with the weather of the upstream and downstream of the sluice and the running condition of the sluice to judge whether the water level before the sluice is stable, and if the water level before the sluice is judged to be stable, the water level before the sluice is directly used as a reference surface for calibration.

As an implementation mode, an underwater tray is used as a reference surface for calibration, after an optimal measurement value error correction model is determined, the underwater tray is erected and folded until the underwater tray is tightly attached to the wall surface where the upper vertical guide rail and the lower vertical guide rail are located, and then the water level before the gate is measured on line.

As an implementation mode, in the calibration process, the water level measuring device and the bracket move up and down periodically within the measuring range of the reflection type water level meter; and when the calibration is finished, the staying and fixing positions of the water level measuring device and the bracket are the middle part of the measuring range of the reflection type water level meter.

As an embodiment, a correlation vector machine (SVM), a long-short time memory (LSTM), an encoding-decoding neural network model, and a gated round robin unit (GRU) are selected as the measured value error correction model.

As an embodiment, the method for comprehensively evaluating the trained measured value error correction model includes: calculating the measurement of errors in a predicted value obtained by measuring the value error correction model by randomly extracting a plurality of groups of input combinations, and judging the precision of the value error correction model; adding a certain error into an input sample, calculating the error in a predicted value obtained by a measured value error correction model, and judging the stability of the measured value error correction model; judging the generalization capability of the measured value error correction model by judging whether the corresponding time length meets the requirement when the prediction accuracy is the same maximum allowable error; and comprehensively considering the model with the best precision, stability and generalization capability as the determined optimal measured value error correction model.

Further, 50 sets of input combinations are randomly extracted, and the measure of the error in the predicted value obtained by measuring the value error correction model is calculated.

Further, a 10% error is added to the input sample, and the absolute value of the change in the predicted value obtained by the measured value error correction model is calculated.

Further, when the prediction precision is less than 2 times of the error, whether the corresponding time length meets the requirement is judged.

Further, the method for comprehensively evaluating the trained measured value error correction model may adopt any one of a fuzzy comprehensive evaluation method and a hierarchical analysis method, and the specific analysis method is as follows: and calculating the precision index, the stability index and the generalization ability index to obtain a comprehensive index, and selecting a total optimal measured value error correction model through sorting according to the comprehensive index obtained by multi-index calculation.

Example 2

As shown in fig. 1-2, the present embodiment provides a system for intelligently calibrating a reflection type water level gauge in front of a gate based on edge calculation, the system comprises a measuring device and a bracket 1, a precise bracket lifting control device 2, a datum plane providing device 3, a measuring communication control device 4, a leveling punctuation 5 and a reinforced concrete pier surface 6, wherein the measuring device and the bracket 1 are used for measuring water level in front of the gate and meteorological and transparency data; the bracket precision lifting control device 2 is used for driving the water level measuring device and the bracket to move up and down at a fixed interval height and to be fixed; the reference surface providing device 3 is used for providing a stable water surface reference during calibration and ensuring that the water flow and the water surface are not influenced during measurement; the measurement communication control device 4 is connected with the measurement device and the support 1, the support precision lifting control device 2 and the reference surface providing device 3, and is used for measuring the parameters of the sensor, controlling the motor, calculating the edge, communicating the front and the back, storing and recording relevant measured data, programs and settings.

Furthermore, the measuring device and the support 1 are vertically liftable components and used as a water level measuring device and a support, and comprise a reflection type water level meter 1-1, a video and water mist concentration probe 1-2, a lifting nut sample driving component 1-3 with internal threads, a cylindrical hollow inner wall smooth positioning sliding suite 1-4, a small meteorological station 1-5 and a cantilever rigid rod piece 1-6; the support precision lifting control device 2 is a vertical lifting driving-guide rod group and comprises a driving motor 2-1, an inclination sensor 2-2, a motor fixing and positioning rigid support 2-3, a positioning smooth guide rod 2-4, a high-precision screw rod 2-5, a vertical bearing 2-6 and an adjustable base 2-7; the reference surface providing device 3 is a reference stable water surface providing device and is used for providing a stable reference water surface when a calibration reference surface cannot be determined, and comprises an underwater tray 3-1, a controllable positioning rotating shaft 3-2, an anchoring rod piece 3-3, a tray up-down lifting guide rail 3-4 and a lifting and control motor 3-5. The measurement communication control device 4 comprises an embedded computer and a power supply communication device.

Further, the reflection type water level meter 1-1 is used for measuring the water surface elevation; the video and water mist concentration probe 1-2 is used for sensing the position of the underwater tray 3-1 and analyzing the air transparency; the lifting nut sample driving component 1-3 with the internal thread is in threaded connection with the high-precision filament stalk 2-5, and the height of the high-precision filament stalk 2-5 is adjusted; a smooth positioning sliding sleeve member 1-4 with a cylindrical hollow inner wall is sleeved on the positioning smooth guide rod 2-4 and slides on the positioning smooth guide rod 2-4; the small meteorological station 1-5 is used for measuring the meteorological factors of the precision and the stability of the reflection type water level meter according to the influence, including wind speed, humidity, air pressure and air temperature data; the cantilever rigid rod piece 1-6 is used for connecting the reflection type water level meter 1-1, the video and water mist concentration probe 1-2, the lifting nut sample driving assembly 1-3 with internal threads, the cylindrical hollow inner wall smooth positioning sliding suite 1-4 and the small weather station 1-5.

Further, the reflection type water level meter 1-1 may be a single ultrasonic wave, a laser, a millimeter wave, or the like, or may be a combination thereof, so as to complement each other to realize mutual connection between high-precision measurement ranges.

Furthermore, a driving motor 2-1 is connected with a high-precision screw rod 2-5, a high-precision stepping motor or a servo motor is adopted for high-precision height control of the water level measuring device and the support, and the device is characterized by small step angle, high positioning precision, no accumulated error, self-locking capability when the device stops running and is not powered off, and reliable operation. The resolution can be determined according to the stepping angle, namely the pulse current number, and the lead of the high-precision screw rod 2-5 after considering temperature expansion and contraction. If the stepping angle is 1.5 degrees, a total of 240 pulse currents are needed when the spindle rotates for one circle, the lead of the high-precision screw rod is 2-5 mm, and the resolution is 0.010417 when the lead of the high-precision screw rod is 2.5mm, and 2.5/240 is the resolution.

Further, the driving motor 2-1 drives the high-precision screw rod 2-5 to rotate forwards and backwards, and the height position of the water level measuring device and the height position of the support are adjusted.

Further, the driving motor 2-1 rotates forward to drive the high-precision screw rod 2-5 to rotate to control the water level measuring device and the support to move upwards, and the driving motor 2-1 rotates backward to drive the high-precision screw rod 2-5 to rotate to control the water level measuring device and the support to move downwards.

Furthermore, the water level measuring device and the bracket move for a period, so that the water level change range before the gate is in the highest precision measuring range of the instrument.

Further, the tilt sensor 2-2 cooperates with an adjustable base 2-7; the motor fixing and positioning rigid support 2-3 is used for fixing the driving motor 2-1; positioning the smooth guide rods 2-4, correcting the height of the guide rods by considering temperature change, and calculating the height of the guide rods by thermal coupling according to the thermal expansion coefficient of the material; the high-precision wire rods 2-5 control the lifting height and determine the specific position of the water level measuring device and the support through the precision threads on the high-precision wire rods, are stainless steel trapezoidal screw rods, and are provided with pressure springs in spiral pairs so as to eliminate gap errors in movement; the vertical bearing 2-6 ensures that the high-precision screw rod 2-5 can freely rotate around a vertical axis, the height of the high-precision screw rod is corrected by considering temperature change, and the high-precision screw rod is obtained by thermal coupling calculation according to the thermal expansion coefficient of a material; the adjustable base 2-7 can check and position the verticality of the smooth guide rod 2-4 according to the inclinometer sensor 2-2.

Furthermore, a guide linear bearing and a sliding support rod are arranged in parallel with the high-precision screw rods 2-5, so that the water level measuring device and the support can be stably and linearly lifted up and down.

Further, the reference surface providing device 3 is used for overcoming natural water level fluctuation and providing a stable reference water surface for the calibration of the reflection type water level gauge. In order to ensure that the water level elevation is unchanged in the calibration sampling process, if the water level of the water gate can provide a stable level in the water gate calibration time, the datum plane providing device 3 is not needed, if the water level fluctuation of the water gate is large, the datum plane providing device 3 is needed, and the manufacturing still water level higher than the front running water level of the water gate is lifted to be used as a datum plane for calibration

Furthermore, the underwater tray 3-1 has two horizontal and vertical positioning states of a water container; during off-line calibration, the underwater tray 3-1 is horizontal, and the underwater tray 3-1 is in a shore-approaching state during ordinary on-line measurement, so that water flow is not influenced. During calibration, the lifting and control motor 3-5 controls the underwater tray 3-1 to be horizontally lifted and filled with water, the position of the underwater tray 3-1 on the tray up-down lifting guide rail 3-4 is adjusted, and the height of the water level of the water is slightly higher than the highest water level coupled by multiple factors such as wind, wave and flow before a gate; during on-line measurement, the lifting and control motor 3-5 controls the underwater tray 3-1 to be tightly attached to the side wall to be upright through the controllable positioning rotating shaft 3-2, so that water flow is prevented from being influenced.

Furthermore, a point is preset on the upper surface of the up-down lifting guide rail 3-4 of the underwater tray for fixed positioning, and the height of the underwater tray 3-1 is lifted and positioned according to the height of the water surface in front of the gate.

Furthermore, all sensors and motors in the device are controlled by the measuring communication control device 4, and form an organic whole under the command of edge calculation.

Furthermore, the embedded computer and the power communication device adopt edge calculation inside and are internally provided with an operating system and software to realize automatic calibration and online measurement of the model. The model adopts a lightweight deep neural network model. The device has the advantages that the device is provided with a measuring device, a control device and a power supply device, can measure the sensors belonging to the whole system, control the motor, supply power, and simultaneously have a remote wireless communication function, can realize two-way communication with a rear management station or a mobile terminal provided with corresponding APP, and realize the operations of parameter setting, sampling control, data reading and writing and the like.

And further, the influence of the temperature on the height and the pitch of the precise screw rod is stored in a local embedded computer according to thermal coupling calculation, and the influence is called according to a corresponding temperature measured value during calibration.

Furthermore, the positioning position and the corresponding height are also stored in the local embedded computer, and the real value which is obtained by the reflection type water level meter, namely the model output value in the calibration stage, is determined by combining the water level measuring device and the lifting height of the bracket during calibration. And in the on-line measurement stage, the output value is the water level value before the gate after error correction.

As an embodiment, the system further comprises leveling points 5 for periodically checking the elevation of the reference surface.

In one embodiment, the system further includes a reinforced concrete pier face 6 that is strong, stable and non-deformable.

In the description herein, references to the description of the terms "one embodiment," "certain embodiments," "an illustrative embodiment," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present invention.