A micro automatic monitoring station for sediment content and a monitoring method

By designing a micro automatic monitoring station for sediment content, using pre-processing systems, analysis units and circular neural network prediction models, real-time monitoring and trend prediction are achieved, and the problem of sediment monitoring in the existing technology is solved. Improve monitoring efficiency and accuracy.

CN118549616BActive Publication Date: 2024-11-22NANJING LIGHT TIMES DIGITAL TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411003337.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-25
Publication Date
2024-11-22
Estimated Expiration
2044-07-25

AI Technical Summary

Technical Problem

The existing sediment monitoring methods take a long time and cannot meet the real-time control needs of sediment concentration, affect subsequent research, and cannot achieve automated and efficient monitoring.

Method used

A micro automatic monitoring station for sediment content was designed, including a pretreatment system, sediment analysis unit, ultrasonic flowmeter/flowmeter and control unit, using multi-neighborhood connected area denoising processing and histogram equalization image processing, combined with Recurrent neural network prediction model realizes real-time monitoring and trend prediction.

Benefits of technology

Automatic monitoring of instantaneous and cumulative sediment content is realized, the efficiency of remote control at the monitoring end is improved, and the accuracy of monitoring results and the long-term automatic operation of the monitoring station is ensured.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118549616B_ABST
    Figure CN118549616B_ABST
Patent Text Reader

Abstract

The present invention discloses a micro automatic monitoring station and a monitoring method for sediment content, and relates to the technical field of sediment monitoring. The monitoring station includes a pretreatment system for water sampling, a sediment analysis unit for detecting and analyzing sediment, an ultrasonic flowmeter / flowmeter for measuring cross-sectional flow, and a control unit, wherein the pretreatment system includes a sampling pump, a sampling tube, an impurity removal device, and a backwashing device, the sampling pump is arranged in a sampling box, and the outlet end of the sampling pump is connected to the pipeline of the sediment analysis unit through the sampling tube; the impurity removal device and the backwashing device are arranged on the sampling tube, and the sampling tube is connected to the impurity removal device; the sediment parameters monitored by the sediment analysis unit include instantaneous sediment content and cumulative sediment content. The present application realizes the full-automatic long-term operation of the monitoring station.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of sediment monitoring, and in particular to a micro automatic monitoring station for sediment content and a monitoring method. Background Art

[0002] In order to effectively monitor the sediment in the water source, so as to timely discover the situation of soil erosion and take targeted measures, frequent sampling and analysis are required. The existing sediment monitoring method mainly collects water samples from the water source, takes a certain amount of river water, dries it, and weighs the mass of the sediment. The ratio of this mass to the volume of water taken is the sediment content of the river water. It can only add an appropriate amount of dust in proportion to the water storage capacity of the riverbed. This method is time-consuming and cannot meet the needs of real-time control of sediment concentration, which seriously affects subsequent research.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] Aiming at the shortcomings of existing sediment monitoring, the present invention discloses a micro automatic monitoring station for sediment content and a monitoring method.

[0005] The present application provides a micro automatic monitoring station for sediment content, comprising a pretreatment system for water sampling, a sediment analysis unit for detecting and analyzing sediment, an ultrasonic flowmeter / velocity meter for measuring cross-sectional flow, and a control unit, wherein the pretreatment system comprises a sampling pump, a sampling tube, an impurity removal device, and a backwashing device, the sampling pump is arranged in a sampling box, and the outlet end of the sampling pump is connected to the pipeline of the sediment analysis unit through the sampling tube; the impurity removal device and the backwashing device are arranged on the sampling tube, and the sampling tube is connected to the impurity removal device; the sediment parameters monitored by the sediment analysis unit include instantaneous sediment content and cumulative sediment content; the control unit is electrically connected to the pretreatment system and the sediment analysis unit.

[0006] Furthermore, the impurity removal device comprises an impurity removal pipeline and an ultrasonic generator arranged on the impurity removal pipeline, wherein the oscillating air generated by the ultrasonic generator is sent to the end of the sampling tube via the impurity removal pipeline; the calculation of the instantaneous sediment content is based on the measurement result of the flow meter / flow meter;

[0007] Furthermore, the backwashing device includes a washing pipeline and a washing water source connected to the washing pipeline, and a valve for switching the communication state is provided between the washing water source and the washing pipeline.

[0008] Furthermore, a microporous isolation grid is provided at the inlet end of the sampling pump, and the pipeline of the sampling tube adopts an algae-removing protective coating material.

[0009] Furthermore, the micro automatic monitoring station for sediment content is also provided with a solar power supply unit, and the solar power supply unit includes a photovoltaic panel and a battery.

[0010] The present application also provides a micro-automatic monitoring method for sediment content, which is based on the above-mentioned micro-automatic monitoring station for sediment content, including: controlling a sampling pump to perform sampling according to a preset sampling and analysis frequency, wherein the sampling pump is arranged in a sampling box, and the outlet end of the sampling pump is connected to the pipeline of a sediment analysis unit through a sampling tube; obtaining sediment parameters by monitoring the water sample sucked by the sampling pump by the sediment analysis unit, wherein the sediment parameters include instantaneous sediment content and cumulative sediment content; a microporous isolation grid is arranged at the inlet end of the sampling pump, and the pipeline of the sampling tube adopts an algae removal and protection coating material; obtaining the cross-sectional flow and cross-sectional flow velocity calculated by an ultrasonic flowmeter based on the instantaneous sediment content; determining the cumulative sediment content and the cumulative sediment content based on the cross-sectional flow and the cross-sectional flow velocity Measuring flow; acquiring multiple images of the target water area taken at different angles; mapping the image of the target water area to a three-dimensional coordinate system by means of three-dimensional image reconstruction, denoising the image of the target water area by using a multi-neighborhood connected region, and performing histogram equalization on the denoised image of the target water area to obtain an enhanced image; predicting the sediment content trend of the target water area by using a sediment content trend prediction model based on the instantaneous sediment content, the cross-sectional flow and the enhanced image, wherein the sediment content trend prediction model is constructed based on a recurrent neural network; storing and uploading the predicted sediment content trend of the target water area, so as to enable the monitoring end to remotely obtain the sediment content trend and remotely control the operation of the sediment content micro-automatic monitoring station.

[0011] Furthermore, the image of the target water area after denoising is subjected to histogram equalization to obtain an enhanced image, including: obtaining an image histogram array according to the grayscale value of each pixel in the image of the target water area after denoising; calculating an equalization mapping curve of the image of the target water area after denoising according to the image histogram array; adding an intensity control factor to the equalization mapping curve to obtain a nonlinear mapping curve; wherein the expression of the nonlinear mapping curve is as follows:

[0012]

[0013] Among them, * is the multiplication sign; is the gray value of each pixel in the image of the target water area after denoising; is the mapping value of the image of the target water area after denoising; P is the intensity control factor; t is the maximum grayscale value among the grayscale values ​​of each pixel in the image of the target water area after denoising; is the total pixel value of the image of the target water area after denoising; is the number of pixels with a gray value of k in the image of the target water area after denoising; is the pixel depth of the image of the target water area after denoising; and performing nonlinear mapping processing on the image of the target water area after denoising according to the nonlinear mapping curve to obtain the enhanced image.

[0014] Furthermore, based on the instantaneous sediment content, the cross-sectional flow and the enhanced image, the sediment content trend of the target water area is predicted using a sediment content trend prediction model, including: the sediment content trend prediction model includes a first prediction model and a second prediction model, the first prediction model is used to make a preliminary prediction of the sediment content trend according to the instantaneous sediment content and the cross-sectional flow, and the second prediction model is used to make a secondary prediction based on the prediction result of the first prediction model and the enhanced image to obtain the sediment content trend; the first prediction model is an extreme gradient boosting model, and the second prediction model is a long-short-term memory model of a memory storage unit; wherein the training error in the objective function of the extreme gradient boosting model is defined as:

[0015] =

[0016] in, It is training data; the training data includes the instantaneous sediment content and the cross-sectional flow; is the total number of training data; is the true label of the training data, is the predicted value given by the t-th tree for the training data, is the predicted value of the training data after the (t-1)th tree, () is a cost value calculated based on the true label and the predicted value, q() is a regular term to prevent the model from overfitting; in the long short-term memory model, the forget gate determines the information to be discarded from the cell state based on the prediction result of the first prediction model and the enhanced image, wherein the cell state is used to control the memory and forgetting of the memory storage unit.

[0017] Furthermore, the instantaneous sediment content and the cumulative sediment content are stored and uploaded, so that the monitoring end can remotely obtain the sediment content monitoring data and remotely control the operation of the monitoring end.

[0018] Furthermore, it is characterized in that the denoising process of the image of the target water area using multi-neighborhood connected regions comprises: using multi-neighborhood connected regions to analyze the image of the target water area, grouping adjacent pixels of the same pixel value into sets, setting a mark for each connected region, and completing connected domain analysis; setting a connected domain area threshold, and deleting connected domains whose pixel area is smaller than the connected domain area threshold.

[0019] The micro automatic monitoring station provided by the present application sets the sampling pump in the sampling box, and the analysis unit analyzes the pre-treated water sample, which is not affected by impurities, ensuring the accuracy of the sampling analysis results; the sampling tube is provided with an impurity removal device and a backwashing device, and the connection state is switched by a valve, which can ensure the cleanliness of the sampling pipeline, so as to improve the accuracy of the sediment analysis results, and can realize the full automatic long-term operation of the monitoring station. The micro automatic monitoring method of sediment content provided by the present application realizes the automatic monitoring of instantaneous sediment content and cumulative sediment content, and the automatic calculation of sediment content trend, which improves the efficiency of remote control of the monitoring station operation by the monitoring end. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings described herein are used to provide a further understanding of the embodiments of the present invention and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0021] Figure 1 is a schematic diagram of an optional micro automatic monitoring station for sediment content according to an embodiment of the present application;

[0022] Figure 2 It is a flow chart of an optional micro-automatic monitoring method of sediment content according to an embodiment of the present application.

[0023] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0024] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0025] The micro automatic monitoring station for sediment content provided by the present invention is described more clearly and completely below by way of a preferred embodiment and in conjunction with the accompanying drawings, but the micro automatic monitoring station for sediment content provided by the present invention is not limited to the scope of the embodiments.

[0026] like Figure 1 The present invention shows a micro automatic monitoring station for sediment content, which includes a pretreatment system for water sampling and a sediment analysis unit 2 for detecting and analyzing sediment. The sediment analysis unit 2 of the present invention monitors the instantaneous sediment content of the sediment parameter, and the cross-sectional flow velocity and flow rate are measured by an ultrasonic flowmeter, and the sediment volume can be accumulated. The pretreatment system includes a sampling pump 3, a sampling tube 4 and a sampling box 5. The sampling pump 3 is arranged in the sampling box 5. The outlet end of the sampling pump 3 is connected to the pipeline of the sediment analysis unit 2 through the sampling tube 4. The sampling pump 3 is used to suck the water sample and send it to the sediment analysis unit 2 for online sediment detection. The pretreatment equipment can be set at the entrance of the sediment analysis unit 2 to ensure that the water sample meets the corresponding instrument test requirements without changing the representativeness of the water sample.

[0027] The sediment analysis unit 2 of the present invention adopts a dual-pump, dual-circuit water collection form, one for use and one for backup, to achieve uninterrupted monitoring, and can automatically or manually realize rotation. The control system is set with automatic diagnosis of pump failure and automatic switching of pump working functions. The water collection system can work in a continuous or intermittent manner, and the monitoring frequency can be set on-site or remotely according to monitoring requirements.

[0028] The sampling tube 4 of the present invention is provided with an impurity removal device (also referred to as an impurity removal facility) and a backwashing device. The sampling tube 4 is connected to the impurity removal device and the backwashing device through a cleaning joint, wherein the backwashing device includes a cleaning pipeline and a cleaning water source connected to the cleaning pipeline, and a valve for switching the connection state is provided between the cleaning water source and the cleaning pipeline, for example, a valve for switching the connection state is provided at the cleaning pipeline joint. When impurities need to be removed, the impurity removal device is connected to the sampling tube 4, and the impurity removal device includes an impurity removal pipeline and an ultrasonic generator provided on the impurity removal pipeline, and the high-frequency oscillating air generated by the ultrasonic generator is sent to the end of the sampling tube, thereby utilizing ultrasonic oxidation to remove impurities; when impurities such as sludge are deposited in the tube due to long-term water sampling, the backwashing device can be connected regularly, and the backwashing device of the present invention includes a cleaning pipeline and a cleaning water source connected to the cleaning pipeline, and the cleaning water source is used to backwash the pipeline to remove the debris in the pipeline, thereby realizing long-term automatic operation of the monitoring station.

[0029] An isolation fence is set at the initial end of the sampling system of the present invention, and an anti-algae coating pipe is set on the sampling pipe 4 for the pipeline. The isolation fence can prevent stones, dead branches, plastic belts, etc. in the water from entering the pipeline and causing damage to the sampling pump. The anti-algae coating pipe prevents algae from being enriched on the pipe wall and affecting the normal operation of the monitoring station.

[0030] A control unit is provided in the monitoring station of the present invention, and the control unit is electrically connected to the pretreatment system and the sediment analysis unit 2. The sampling pump is controlled to perform sampling according to a preset sampling and analysis frequency, and the control system of the sediment analysis unit calculates, stores and uploads the detection result of the sediment analysis unit, thereby realizing remote acquisition of monitoring data and control of the operation of the monitoring station. The monitoring station of the present invention is provided with a solar power supply unit, which includes a photovoltaic panel and a battery. The equipment can be powered by solar energy to adapt to the operation requirements in various environments.

[0031] It should be understood that the monitoring station of the present invention is provided with a solar power supply unit and stores electric energy through batteries, so that when there is a problem with the mains power supply, the normal operation of the monitoring station can still be maintained.

[0032] When the automatic sediment content monitoring station provided by the present invention is in use, the control unit controls the sampling pump 3 to suck water samples into the monitoring station, and the sediment analysis equipment in the sediment analysis unit analyzes the collected water samples to obtain corresponding detection parameters, which are collected, stored and uploaded to the background by the control unit so that the corresponding data can be retrieved at any time. When sampling, the sampling tube is decontaminated and backwashed according to a preset cycle to keep the sampling tube clean.

[0033] The micro automatic monitoring station for sediment content provided by the present invention can be used by floating in water, and can freely detect sediment contents at different depths by changing the style of the carrier.

[0034] The micro automatic monitoring station provided in the present application sets the sampling pump in the sampling box, and the analysis unit analyzes the pretreated water sample without being affected by impurities, thereby ensuring the accuracy of the sampling and analysis results; an impurity removal device and a backwashing device are set on the sampling pipe, and the connection state is switched by a valve to ensure the cleanliness of the sampling pipe, so as to improve the accuracy of the sediment analysis results and realize the fully automatic long-term operation of the monitoring station.

[0035] like Figure 2 As shown, the present application provides a micro automatic monitoring method for sediment content, based on the above-mentioned micro automatic monitoring station for sediment content, comprising:

[0036] S201, controlling a sampling pump to perform sampling according to a preset sampling and analysis frequency, wherein the sampling pump is disposed in a sampling box, and an outlet end of the sampling pump is connected to a pipeline of a sediment analysis unit through a sampling tube;

[0037] S202, obtaining sediment analysis unit to monitor the water sample sucked by the sampling pump to obtain sediment parameters, wherein the sediment parameters include instantaneous sediment content and cumulative sediment content; a microporous isolation grid is set at the inlet end of the sampling pump, and the pipeline of the sampling tube adopts algae removal protection coating material;

[0038] S203, obtaining the cross-sectional flow rate and cross-sectional flow velocity calculated by the ultrasonic flow and velocity meter based on the instantaneous sediment content; determining the cumulative sediment content and the cumulative flow rate based on the cross-sectional flow rate and the cross-sectional flow velocity;

[0039] S204, acquiring multiple images of the target water area taken at different angles; mapping the image of the target water area to a three-dimensional coordinate system by means of three-dimensional image reconstruction, performing denoising on the image of the target water area by using a multi-neighborhood connected region, and performing histogram equalization on the denoised image of the target water area to obtain an enhanced image;

[0040] S205, predicting the sediment content trend of the target water area using a sediment content trend prediction model based on the instantaneous sediment content, cross-sectional flow and enhanced image, wherein the sediment content trend prediction model is constructed based on a recurrent neural network;

[0041] S206, storing and uploading the predicted sediment content trend of the target water area, so that the monitoring end can remotely obtain the sediment content trend and remotely control the operation of the sediment content micro-automatic monitoring station.

[0042] The micro-automatic sediment content monitoring method provided in the present application realizes automatic monitoring of instantaneous sediment content and cumulative sediment content, and automatic calculation of sediment content trend, thereby improving the efficiency of remote control of the monitoring station operation by the monitoring end.

[0043] Furthermore, the image of the target water area after denoising is subjected to histogram equalization to obtain an enhanced image, including:

[0044] According to the gray value of each pixel in the image of the target water area after denoising, an image histogram array is obtained;

[0045] An equalization mapping curve of the image of the target water area after denoising is calculated according to the image histogram array;

[0046] The intensity control factor is added to the equalization mapping curve to obtain a nonlinear mapping curve; wherein, the expression of the nonlinear mapping curve is as follows:

[0047]

[0048] Among them, * is the multiplication sign; is the gray value of each pixel in the image of the target water area after denoising; is the mapping value of the image of the target water area after denoising; P is the intensity control factor; t is the maximum grayscale value among the grayscale values ​​of each pixel in the image of the target water area after denoising; is the total pixel value of the target water area image after denoising; is the number of pixels with gray value k in the image of the target water area after denoising; is the pixel depth of the target water area image after denoising;

[0049] The image of the target water area after denoising is subjected to nonlinear mapping processing according to the nonlinear mapping curve to obtain an enhanced image.

[0050] Furthermore, based on the instantaneous sediment content, cross-sectional flow and enhanced images, the sediment content trend prediction model is used to predict the sediment content trend of the target waters, including:

[0051] The sediment content trend prediction model includes a first prediction model and a second prediction model. The first prediction model is used to make a preliminary prediction of the sediment content trend according to the instantaneous sediment content and the cross-sectional flow rate. The second prediction model is used to make a secondary prediction according to the prediction result of the first prediction model and the enhanced image to obtain the sediment content trend.

[0052] The first prediction model is an extreme gradient boosting model, and the second prediction model is a long short-term memory model of a memory storage unit; wherein the training error in the objective function of the extreme gradient boosting model is defined as:

[0053] =

[0054] in, It is training data; the training data includes instantaneous sediment content and cross-sectional flow; is the total number of training data; is the true label of the training data, is the prediction value given by the (t-1)th tree for the training data, is the predicted value of the training data after the tth tree, () is the cost value calculated based on the true label and the predicted value, and q() is a regular term to prevent the model from overfitting;

[0055] In the long short-term memory model, the forget gate determines the information to be discarded from the cell state based on the prediction result of the first prediction model and the enhanced image, wherein the cell state is used to control the memory and forgetting of the memory storage unit.

[0056] Furthermore, the instantaneous sediment content and the cumulative sediment content are stored and uploaded, so that the monitoring end can remotely obtain the sediment content monitoring data and remotely control the operation of the monitoring end.

[0057] Furthermore, it is characterized in that the image of the target water area is denoised using multi-neighborhood connected regions, including:

[0058] The multi-neighborhood connected region is used to analyze the image of the target water area. Adjacent pixels with the same pixel value are grouped into sets, and a mark is set for each connected region to complete the connected domain analysis. A connected domain area threshold is set, and connected domains with pixel areas smaller than the connected domain area threshold are deleted.

[0059] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A micro automatic monitoring method for sediment content, based on a micro automatic monitoring station for sediment content, the automatic monitoring station includes a pretreatment system for water sampling, a sediment analysis unit for detecting and analyzing sediment, an ultrasonic flowmeter / flowmeter for measuring cross-sectional flow, and a control unit, wherein: The pretreatment system comprises a sampling pump, a sampling tube, an impurity removal device and a backwashing device, wherein the sampling pump is arranged in a sampling box, and the outlet end of the sampling pump is connected to the pipeline of the sediment analysis unit through the sampling tube; The impurity removal device and the backwashing device are arranged on the sampling tube, and the sampling tube is connected to the impurity removal device; The sediment parameters monitored by the sediment analysis unit include instantaneous sediment content and cumulative sediment content; The control unit is electrically connected to the pre-processing system and the sediment analysis unit; The invention is characterized by comprising: Controlling a sampling pump to perform sampling according to a preset sampling and analysis frequency, wherein the sampling pump is arranged in a sampling box, and an outlet end of the sampling pump is connected to a pipeline of a sediment analysis unit through a sampling pipe; The sediment analysis unit is used to monitor the water sample sucked by the sampling pump to obtain sediment parameters, wherein the sediment parameters include instantaneous sediment content and cumulative sediment content; a microporous isolation grid is provided at the inlet end of the sampling pump, and the pipeline of the sampling tube adopts an algae removal and protection coating material; Obtaining the cross-sectional flow rate and cross-sectional flow velocity calculated by the ultrasonic flow and velocity meter based on the instantaneous sediment content; determining the cumulative sediment content and the cumulative flow rate based on the cross-sectional flow rate and the cross-sectional flow velocity; Acquire multiple images of the target water area taken at different angles; map the image of the target water area to a three-dimensional coordinate system by means of three-dimensional image reconstruction, perform denoising on the image of the target water area by using a multi-neighborhood connected region, and perform histogram equalization on the denoised image of the target water area to obtain an enhanced image; Based on the instantaneous sediment content, the cross-sectional flow and the enhanced image, a sediment content trend prediction model is used to predict the sediment content trend of the target water area, wherein the sediment content trend prediction model is constructed based on a recurrent neural network; The predicted sediment content trend of the target water area is stored and uploaded, so that the monitoring end can remotely obtain the sediment content trend and remotely control the operation of the sediment content micro-automatic monitoring station; The image of the target water area after denoising is subjected to histogram equalization to obtain an enhanced image, including: Obtaining an image histogram array according to the grayscale value of each pixel in the image of the target water area after denoising; Calculating an equalization mapping curve of the image of the target water area after denoising according to the image histogram array; A strength control factor is added to the equalization mapping curve to obtain a nonlinear mapping curve; wherein the expression of the nonlinear mapping curve is as follows: Among them, * is the multiplication sign; is the gray value of each pixel in the image of the target water area after denoising; is the mapping value of the image of the target water area after denoising; P is the intensity control factor; t is the maximum grayscale value among the grayscale values ​​of each pixel in the image of the target water area after denoising; is the total pixel value of the image of the target water area after denoising; is the number of pixels with a gray value of k in the image of the target water area after denoising; is the pixel depth of the image of the target water area after denoising; Performing nonlinear mapping processing on the denoised image of the target water area according to the nonlinear mapping curve to obtain the enhanced image; The method of predicting the sediment content trend of the target water area by using a sediment content trend prediction model based on the instantaneous sediment content, the cross-sectional flow and the enhanced image includes: The sediment content trend prediction model includes a first prediction model and a second prediction model, wherein the first prediction model is used to make a preliminary prediction of the sediment content trend according to the instantaneous sediment content and the cross-sectional flow, and the second prediction model is used to make a secondary prediction according to the prediction result of the first prediction model and the enhanced image to obtain the sediment content trend; The first prediction model is an extreme gradient boosting model, and the second prediction model is a long short-term memory model of a memory storage unit; wherein the training error in the objective function of the extreme gradient boosting model is defined as: = in, It is training data; the training data includes the instantaneous sediment content and the cross-sectional flow; is the total number of training data; is the true label of the training data, is the predicted value given by the t-th tree for the training data, is the predicted value of the training data after the (t-1)th tree, () is the cost value calculated based on the true label and the predicted value, and q() is a regularization term to prevent the model from overfitting; In the long short-term memory model, a forget gate determines information discarded from a cell state based on a prediction result of the first prediction model and the enhanced image, wherein the cell state is used to control memory and forgetting of the memory storage unit; The denoising process of the image of the target water area by using multi-neighborhood connected regions includes: The image of the target water area is analyzed using a multi-neighborhood connected region, adjacent pixels with the same pixel value are grouped into a set, a mark is set for each connected region, and a connected domain analysis is completed; a connected domain area threshold is set, and connected domains with pixel areas smaller than the connected domain area threshold are deleted.

2. The micro automatic monitoring method for sediment content according to claim 1 is characterized in that: The impurity removal device comprises an impurity removal pipeline and an ultrasonic generator arranged on the impurity removal pipeline, wherein the oscillating air generated by the ultrasonic generator is sent to the end of the sampling tube via the impurity removal pipeline; The calculation of the instantaneous sediment content is based on the measurement results of the flow meter / velocity meter.

3. The micro automatic monitoring method for sediment content according to claim 2 is characterized in that: The backwashing device comprises a washing pipeline and a washing water source connected to the washing pipeline, and a valve for switching the connection state is arranged between the washing water source and the washing pipeline.

4. The micro automatic monitoring method for sediment content according to claim 3 is characterized in that: A microporous isolation grid is arranged at the inlet end of the sampling pump, and the pipeline of the sampling tube adopts an algae-removing protective coating material.

5. The micro automatic monitoring method for sediment content according to any one of claims 1 to 4, characterized in that: The micro automatic monitoring station for sediment content is also provided with a solar power supply unit, which includes a photovoltaic panel and a storage battery.

6. The micro automatic monitoring method for sediment content according to claim 5 is characterized in that: The instantaneous sediment content and the cumulative sediment content are stored and uploaded, so that the monitoring end can remotely obtain the sediment content monitoring data and remotely control the operation of the monitoring end.