Water pump water pumping automatic adjusting method and system for preventing river channel mud bottom disturbance

Abstract

The invention relates to the field of automatic riverway water pumping, in particular to a water pump water pumping automatic adjusting method and system for preventing riverway mud bottom disturbance. The method comprises the following steps: acquiring water pumping data related to mud bottom disturbance, and evaluating the influence of water pumping on river sediment according to the water pumping data to obtain a disturbance degree; obtaining a sample set according to the water pumping data and the disturbance degree; constructing a water pump water inlet prediction model, and setting a loss function for the water pump water inlet prediction model according to the disturbance degree; training and evaluating the water pump water inlet prediction model by using the sample set; and predicting by using the water pump water inlet prediction model, and automatically adjusting the water inlet and the power of the water pump according to the prediction result. The problem that in the prior art, sediment disturbance cannot be effectively controlled during water pumping, and consequently heavy metal pollution of farmland is caused is solved. Automatic water pumping is achieved, and the water pumping efficiency is improved through a scientific method.

Description

Technical Field

[0001] The invention relates to the field of automatic water pumping in rivers, and in particular to an automatic water pumping adjustment method and system for preventing muddy bottom disturbance in rivers. Background Art

[0002] In agricultural production, river water is one of the important irrigation water sources. The water quality of many rivers can meet the requirements of the "Agricultural Irrigation Water Quality Standard" (GB 5084-2021), providing the necessary water support for crop growth and ensuring the basic production activities of agriculture. However, the problem of river pollution cannot be ignored, especially heavy metal pollution. After entering the river through wastewater discharge, soil erosion, rainwater erosion and other channels, heavy metals will gradually accumulate in the sediment through precipitation or adsorption of particulate matter. This has created a situation where the river water can meet the agricultural irrigation water quality standards, but the heavy metals in the river sediment have seriously exceeded the "Soil Environmental Quality Agricultural Land Soil Pollution Risk Control Standard (Trial)" (GB 15618-2018). Sediment pollution is a relatively common environmental problem and one of the main difficulties in the process of river and lake pollution control. The contaminated sediment is large in volume, wide in area, high in cost, difficult to control, and it is impossible to detect or control it in time. Therefore, it is particularly important to prevent the contaminated sediment from entering farmland.

[0003] When pumping water from rivers to irrigate farmland, a key problem surfaces, namely how to prevent sediment disturbance. On the one hand, heavy metals enriched in the sediment are more likely to re-enter the water body once disturbed. On the other hand, the disturbed sediment will be pumped out with the irrigation water and enter the farmland during the irrigation process. The existing technology is difficult to effectively control sediment disturbance during the pumping process. Common pumping equipment, such as water pumps, will produce suction and stirring effects when working. When the water inlet of the water pump is close to the sediment, or the pumping flow is too large and the flow rate is too fast, the impact force of the water flow will drive the sediment and suspend it in the water.

[0004] In addition, some traditional pumping methods lack targeted designs for bottom sediment protection and do not fully consider the potential hazards caused by bottom sediment disturbance. For example, when installing pumping equipment at the bottom of the river, no reasonable protective measures are set, which inevitably causes the bottom sediment to be stirred during the pumping process.

[0005] The consequences of this sediment disturbance are very serious. When irrigation water containing contaminated sediment enters farmland, heavy metals will gradually accumulate in the soil, changing the physical and chemical properties of the soil and reducing soil fertility. At the same time, crops will absorb heavy metals in the soil, causing excessive heavy metal content in agricultural products, which not only affects the growth and quality of crops, but also harms human health through the food chain.

[0006] Therefore, it is urgent to develop a technology that can effectively control sediment disturbance during river pumping irrigation, which is of great significance for ensuring the soil quality of farmland and the safety of agricultural products and achieving sustainable agricultural development. Summary of the invention

[0007] In view of the defects in the prior art, the present invention provides a method and system for automatically adjusting water pumping for preventing mud disturbance in the riverbed, thereby solving the problem that the prior art cannot effectively control mud disturbance during pumping, thereby causing heavy metal pollution in farmland.

[0008] In order to achieve the above-mentioned purpose, one aspect of the present invention provides a method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river channel, the method comprising: obtaining pumping data related to muddy bottom disturbance, and evaluating the impact of pumping on river channel bottom mud based on the pumping data to obtain the degree of disturbance; obtaining a sample set based on the pumping data and the degree of disturbance; constructing a water pump water inlet prediction model, and setting a loss function for the water pump water inlet prediction model based on the degree of disturbance; using the sample set to train and evaluate the water pump water inlet prediction model; using the water pump water inlet prediction model to make predictions, and automatically adjusting the water inlet and power of the water pump based on the prediction results.

[0009] The present invention obtains pumping data related to mud bottom disturbance, evaluates the impact of pumping on riverbed mud to obtain the degree of disturbance, and then constructs a sample set, providing a key data basis for subsequent research. A water pump inlet prediction model is constructed and a loss function is set according to the degree of disturbance, so that the model can more accurately reflect the actual situation and improve the prediction accuracy. The trained model is used for prediction, and the water pump inlet and power are automatically adjusted accordingly, so that the water pump inlet and power can be dynamically adjusted while preventing mud disturbance.

[0010] Optionally, the obtaining of pumping data related to mud bottom disturbance includes: conducting a pumping test and testing the riverbed mud to obtain initial data; performing correlation analysis on the initial data to obtain core parameters of mud disturbance; and using the core parameters of mud disturbance to screen the initial data to obtain the pumping data related to mud bottom disturbance.

[0011] The present invention obtains the core parameters of sediment disturbance by performing correlation analysis on the initial data, accurately extracts the factors that play a key role in the mud bottom disturbance, helps to focus on key research directions, avoids blind exploration in massive data, reduces the amount of data calculation, and makes the present invention more scientific, accurate and efficient in actual application.

[0012] Optionally, performing correlation analysis on the initial data to obtain the core parameters of sediment disturbance includes: obtaining sediment disturbance parameters based on the initial data; calculating the correlation coefficients between the sediment disturbance parameters, and constructing a theoretical correlation coefficient matrix based on the correlation coefficients; performing eigenvalue decomposition on the theoretical correlation coefficient matrix to obtain eigenvectors and eigenvalues; calculating the contribution of the sediment disturbance parameters to the eigenvectors based on the eigenvectors and the eigenvalues; and screening the sediment disturbance parameters using the contribution to obtain the core parameters of sediment disturbance.

[0013] The present invention can intuitively show the degree of correlation between parameters by calculating the correlation coefficient and constructing a theoretical correlation coefficient matrix, which helps to gain insight into the internal structure of the data. The eigenvalue decomposition is used to obtain the eigenvector and eigenvalue, which effectively mines the potential information of the data. By calculating the contribution of the sediment disturbance parameters on the eigenvector, the influence of each parameter on the overall disturbance can be quantified, which realizes the accurate extraction of key information from complex data and improves the accuracy of screening the core parameters of sediment disturbance.

[0014] Optionally, obtaining the sample set based on the pumping data and the disturbance degree includes: eliminating outliers and interpolating missing values ​​from the pumping data and the disturbance degree to obtain first optimized data; normalizing the first optimized data to obtain second optimized data; constructing an adversarial network model, and generating enhanced data based on the second optimized data and the adversarial network model; and merging the second optimized data and the enhanced data to obtain the sample set.

[0015] The present invention can effectively purify the data by removing outliers and interpolating missing values ​​from the pumping data and the degree of disturbance, avoid misleading subsequent analysis caused by abnormal data, fill in missing information, ensure the integrity and reliability of the data, and provide a solid data foundation for research. The normalization process converts the first optimized data into a unified scale, so that different features are in a comparable range, which helps to improve the model convergence speed and prediction accuracy, and avoid model training deviations due to differences in data scales. Constructing an adversarial network model to generate enhanced data further expands the diversity of samples, alleviates the problem of insufficient or uneven sample size, and reduces the cost of pumping tests.

[0016] Optionally, setting a loss function for the water pump inlet prediction model according to the disturbance degree includes: based on the pumping test, sampling and testing the heavy metal content at the outlet of the water pump; based on the results of the sampling test, evaluating the disturbance of the riverbed sediment to obtain the maximum acceptable sediment disturbance degree when the water pump is pumping water; and setting a loss function for the water pump inlet prediction model according to the maximum sediment disturbance degree and the disturbance degree.

[0017] The present invention detects the heavy metal content of the water outlet of the water pump according to the maximum acceptable heavy metal content of the farmland, determines the pumping data corresponding to the maximum acceptable heavy metal content, and further calculates the maximum sediment disturbance degree corresponding to the pumping data, and sets the loss function according to the maximum sediment disturbance degree and the actual disturbance degree, so that the water pump inlet prediction model can be closely combined with the actual environmental requirements. During the training process of the water pump inlet prediction model, the loss function guides the model to optimize in the direction of reducing the deviation between the actual disturbance degree and the maximum acceptable disturbance degree, so as to make the prediction results of the water pump inlet prediction model more in line with the ecological safety standards, accelerate the training speed of the water pump inlet prediction model, and improve the prediction effect of the water pump inlet prediction model.

[0018] Optionally, the prediction using the water pump water inlet prediction model and automatically adjusting the water pump water inlet and power according to the prediction result includes: obtaining real-time monitoring data, and predicting using the water pump water inlet prediction model based on the real-time monitoring data; obtaining the real-time distance from the water inlet to the water surface, and automatically adjusting the water pump water inlet and power according to the real-time distance and the prediction result.

[0019] The present invention can obtain real-time monitoring data and make predictions with the help of prediction models, so as to grasp the relevant status and trend of the water pump in real time during operation, and provide accurate basis for subsequent adjustment. Combined with the distance information from the water inlet to the water surface, it can realize more scientific and reasonable automatic adjustment of the water inlet and power of the water pump, thus improving the operation efficiency of the water pump.

[0020] Optionally, the automatic adjustment of the water inlet and power of the water pump according to the real-time distance and the predicted result includes: according to the working condition of the water pump, setting the minimum distance from the water inlet of the water pump to the water surface when the water pump is working normally as a threshold; and automatically adjusting the water inlet and power of the water pump according to the distance and the threshold.

[0021] The present invention incorporates the minimum distance from the water inlet to the water surface when the water pump can work normally into the adjustment basis of the water inlet, thereby avoiding cavitation and other faults caused by the water inlet being too close to the water surface. At the same time, it can also avoid the water inlet jumping out of the water surface under the action of automatic adjustment, causing invalid work, thereby extending the service life of the water pump, reducing maintenance costs, and improving the feasibility and applicability of the present invention.

[0022] Optionally, the disturbance degree satisfies the following formula: , in, is the degree of mud bottom disturbance, is the comprehensive correction factor, is the pumping power of the water pump, is the distance from the water pump inlet to the mud bottom, is the river flow velocity, is the acceleration due to gravity, is the average diameter of mud bottom particles, is the density of the mud bottom, is the density of river water, is the water flow velocity at the water inlet.

[0023] The disturbance degree formula of the present invention comprehensively considers many key factors such as the water pumping power, the distance from the water inlet to the mud bottom, the water flow velocity of the river channel, etc., thereby improving the accuracy of the calculation of the sediment disturbance degree.

[0024] Optionally, the loss function satisfies the following formula: , in, is the loss function, is the sample size, For the The degree of sediment disturbance of each sample, The maximum acceptable level of sediment disturbance when pumping water.

[0025] The present invention provides a clear optimization direction for the training of the water pump inlet prediction model by calculating the difference between the sample sediment disturbance degree and the maximum acceptable sediment disturbance degree. At the same time, the formula structure is simple and easy to calculate.

[0026] Another aspect of the present invention provides a water pump automatic regulation system for preventing muddy bottom disturbance in a river channel, the system comprising: a processor, an input device, an output device and a memory, the processor, the input device, the output device and the memory being interconnected, wherein the memory is used to store a computer program, the computer program comprising program instructions, the processor being configured to call the program instructions to execute any one of the water pump automatic regulation methods for preventing muddy bottom disturbance in a river channel provided in the previous aspect of the present invention.

[0027] The automatic regulating system for water pumping for preventing muddy bottom disturbance in a river channel of the present invention has a compact structure, stable performance, high integration and simple composition, and can stably execute the automatic regulating method for water pumping for preventing muddy bottom disturbance in a river channel provided in the previous aspect of the present invention, thereby further improving the overall applicability and practical application capability of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A flow chart of a method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river channel according to an embodiment of the present invention; Figure 2 The present invention is a schematic diagram of the structure of an automatic water pumping adjustment system for preventing muddy bottom disturbance in a river channel according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The specific embodiments of the present invention will be described in detail below. It should be noted that the embodiments described herein are only for illustration and are not intended to limit the present invention. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present invention. However, it is obvious to those of ordinary skill in the art that these specific details do not need to be adopted to implement the present invention. In other examples, in order to avoid confusing the present invention, known circuits, software or methods are not specifically described.

[0030] Throughout the specification, references to "one embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in conjunction with the embodiment or example is included in at least one embodiment of the present invention. Therefore, the phrases "in one embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily all refer to the same embodiment or example. In addition, particular features, structures, or characteristics may be combined in one or more embodiments or examples in any suitable combination and / or subcombination. In addition, it should be understood by those of ordinary skill in the art that the figures provided herein are for illustrative purposes and that the figures are not necessarily drawn to scale.

[0031] Figure 1 The flowchart of the automatic adjustment method of a water pump to prevent the disturbance of the muddy bottom of a river channel according to an embodiment of the present invention is shown in FIG. In order to solve the problem that the disturbance of the muddy bottom cannot be effectively controlled during pumping in the prior art, thus causing heavy metal pollution in farmland, such as Figure 1 The method shown includes the following steps: Step S1, obtaining pumping data related to mud bottom disturbance, and evaluating the impact of pumping on riverbed mud based on the pumping data to obtain the degree of disturbance.

[0032] The step of obtaining pumping data related to mud bottom disturbance specifically includes the following sub-steps: Step S101, conduct a pumping test and detect the riverbed mud to obtain initial data.

[0033] In this embodiment, in order to ensure the scientificity and efficiency of the pumping operation, it is necessary to carefully determine the key parameters of the initial data based on empirical methods and data investigation. These parameters include river flow velocity, river water temperature, river water density, sediment density, average sediment particle size, pump power, and pump inlet diameter. Each parameter plays a vital role in the planning and implementation of subsequent pumping operations. They are interrelated and jointly affect the effect and efficiency of pumping.

[0034] At present, the water demand for farmland needs to be alleviated urgently. In order to meet this urgent demand, it is necessary to carefully plan the water diversion channel according to the distribution of farmland and its relative position to the river. In the process of planning the water diversion channel, comprehensive and meticulous preliminary preparation is essential. For example, a detailed survey of the selected extraction point is carried out. Understanding the geological structure helps to determine the bearing capacity of the soil and its impact on the stability of the water pump installation; the clear distribution of the aquifer is related to the sustainability of the water source and the water supply; the mastery of the river topography can help planners to reasonably layout the water pumps and pipelines to avoid poor water flow or equipment damage due to factors such as terrain undulations.

[0035] At the same time, a preliminary assessment of the hydrogeological conditions at the extraction point is also a key link. The height of the water level determines the installation height and pumping head of the water pump. The quality of water directly affects the quality of agricultural irrigation and the service life of the water pump. The determination of the direction of water flow helps to reasonably plan the direction of the water diversion channel to ensure that the water source can be smoothly introduced into the farmland. Through the comprehensive consideration and analysis of these factors, a solid foundation is laid for the subsequent stable and efficient pumping operation.

[0036] After in-depth investigation and scientific demonstration by relevant units, and taking into full consideration the actual water demand of local farmland, it was finally decided to carefully select 4 most suitable pumping points from the many selected pumping points to install water pumps. After multiple comparisons and considerations, the centrifugal pump model IS200-150-315 was selected. This type of centrifugal pump has excellent performance, a maximum power of 110 kilowatts, and can provide strong power support. Its water inlet diameter is 0.2 meters, the flow range is between 12.5 cubic meters / hour and 400 cubic meters / hour, and the head range is between 8 meters and 125 meters. The pumping parameters can be flexibly adjusted according to different irrigation needs, fully meeting the diverse water requirements of farmland.

[0037] After selecting the water pump, connecting the pumping pipe system becomes an important task. The system includes a suction pipe and a discharge pipe, each of which performs its own function. The suction pipe must have good sealing properties to prevent air from entering and affecting the water absorption efficiency. At the same time, the pipe diameter must be large enough to ensure the smooth suction of water. The discharge pipe must be reasonably set to an appropriate length and diameter based on the actual test requirements to ensure that the water can be discharged without hindrance and smoothly transported to the farmland. In addition, necessary valves and connectors must be installed throughout the entire pipeline system. Valves are used to flexibly control the on-off and flow rate of water flow, which is convenient for adjustment according to actual irrigation conditions. Connectors ensure the stable connection of the pipeline system, facilitate subsequent equipment maintenance and overhaul work, and ensure that the entire pumping system can operate stably and for a long time.

[0038] After all the installation work is completed, a pumping test is carried out. By adjusting the different positions of the water inlet, different pumping parameters are obtained. The river data is tested. The propeller current meter is used to measure the speed of the propeller and convert the water flow velocity into the relevant formula. The temperature of the river water is measured by an electronic thermometer, the density of the river water is measured by a densitometer, the density of the river bottom is measured by the ring knife method, and the average diameter of the mud bottom particles is measured by a laser particle size analyzer.

[0039] Finally, all parameters are integrated to form initial data.

[0040] Step S102, performing correlation analysis on the initial data to obtain core parameters of sediment disturbance.

[0041] The step of performing correlation analysis on the initial data to obtain the core parameters of sediment disturbance specifically includes the following sub-steps: Step S10201, obtaining sediment disturbance parameters according to the initial data.

[0042] In this embodiment, the initial data are distinguished according to their types, each type corresponds to a parameter, and the obtained parameters are then integrated to form sediment disturbance parameters.

[0043] Step S10202, calculate the correlation coefficients between the sediment disturbance parameters, and construct a theoretical correlation coefficient matrix based on the correlation coefficients.

[0044] In this embodiment, the correlation coefficients between the sediment disturbance parameters are first calculated to obtain the correlation coefficients, and a theoretical correlation coefficient matrix is ​​constructed based on the correlation coefficients.

[0045] The correlation coefficient satisfies the following formula: , in, is the sediment disturbance parameter and sediment disturbance parameters The correlation coefficient between is the number of observations of sediment disturbance parameters, is the sediment disturbance parameter No. Observations, is the sediment disturbance parameter The average value of is the sediment disturbance parameter No. Observations, is the sediment disturbance parameter The average value of .

[0046] The correlation coefficient reflects the strength and direction of the linear relationship between two sediment disturbance parameter variables, and its value range is If the correlation coefficient is close to 1, it means that the two sediment disturbance parameters are strongly positively correlated; if the correlation coefficient is close to -1, it is strongly negatively correlated; if the correlation coefficient is close to 0, it means that the linear relationship between the two sediment disturbance parameter variables is weak.

[0047] According to the calculated correlation coefficient, the correlation coefficients of each sediment disturbance parameter and other sediment disturbance parameters are arranged in sequence to construct a theoretical correlation coefficient matrix. The main diagonal elements of the theoretical correlation coefficient matrix are usually 1, representing the correlation between the sediment disturbance parameter itself and the sediment disturbance parameter. This matrix comprehensively displays the degree of correlation between all sediment disturbance parameters in an intuitive and orderly manner.

[0048] Step S10203, performing eigenvalue decomposition on the theoretical correlation coefficient matrix to obtain eigenvectors and eigenvalues.

[0049] In this embodiment, eigenvalue decomposition is an important matrix operation method. Its principle is to decompose the theoretical correlation coefficient matrix into the product of three matrices through a specific mathematical algorithm: a matrix composed of eigenvectors, a diagonal matrix whose main diagonal elements are eigenvalues, and an inverse matrix of the eigenvector matrix.

[0050] In an optional embodiment, the theoretical correlation coefficient matrix is ​​input into MATLAB, which is a powerful mathematical computing software. MATLAB uses the eig function to perform eigenvalue decomposition. The eig function is a tool specifically used to perform this operation. When the eig function receives the theoretical correlation coefficient matrix, the eig function will operate the matrix according to a specific mathematical algorithm. During the operation, MATLAB will look for vectors that make the product of the theoretical correlation coefficient matrix and the vector equal to the vector multiplied by the eigenvalue, thereby outputting two results, one of which is a matrix composed of eigenvectors, in which each column is an eigenvector corresponding to a certain eigenvalue. These eigenvectors reflect the change pattern of data in different directions. The other result is a diagonal matrix, and the elements on its main diagonal are the corresponding eigenvalues. The size of the eigenvalue indicates the degree of data change in the direction of the corresponding eigenvector, and a larger eigenvalue means that the data has a more significant change in this direction.

[0051] Step S10204: Calculate the contribution of the sediment disturbance parameter to the eigenvector based on the eigenvector and the eigenvalue.

[0052] In this embodiment, since the eigenvector is obtained by performing eigenvalue decomposition on the theoretical correlation coefficient matrix, the eigenvector represents the comprehensive change direction of the sediment disturbance parameter in different dimensions. Each eigenvector is composed of multiple components, which correspond to various sediment disturbance parameters, and the eigenvalue reflects the importance of the information contained in the corresponding eigenvector.

[0053] Therefore, the square of the eigenvector component corresponding to each sediment disturbance parameter is multiplied by the eigenvalue corresponding to the eigenvector to obtain the weighted square value of the component. Then, the sum of the square values ​​of the weighted components corresponding to all sediment disturbance parameters is calculated. Finally, the square value of the weighted component corresponding to each sediment disturbance parameter is divided by the sum of the square values ​​of the weighted components to quantify the contribution ratio of the sediment disturbance parameter in the direction of a specific eigenvector. Through this method, a clear contribution value can be obtained for each sediment disturbance parameter and each eigenvector.

[0054] By calculating the contribution, the embodiment of the present invention can clearly understand the relative importance of each sediment disturbance parameter in different eigenvector directions. For example, if a sediment disturbance parameter has a large contribution on a certain eigenvector, it means that the parameter plays a key role in the sediment disturbance in this specific comprehensive change direction. This helps to quickly locate the parameter combination and direction that has the most significant impact on the sediment disturbance among many sediment disturbance parameters, thereby providing a strong basis for more accurate screening of the core parameters of sediment disturbance in the future.

[0055] Step S10205, using the contribution degree to screen the sediment disturbance parameters to obtain the sediment disturbance core parameters.

[0056] In this embodiment, a reasonable contribution threshold is set based on experience or multiple tests, and the contribution of each sediment disturbance parameter is compared with it. The parameter with a contribution greater than the threshold is identified as the core parameter of sediment disturbance. The screening can avoid the complexity and interference caused by too many sediment disturbance parameters, which not only improves the research efficiency, but also can more accurately grasp the essence of sediment disturbance.

[0057] The disturbance degree satisfies the following formula: , in, is the degree of mud bottom disturbance, is the comprehensive correction factor, is the pumping power of the water pump, is the distance from the water pump inlet to the mud bottom, is the river flow velocity, is the acceleration due to gravity, is the average diameter of mud bottom particles, is the density of the mud bottom, is the density of river water, is the water flow velocity at the water inlet.

[0058] The water flow velocity at the water inlet satisfies the following formula: , in, is the water flow velocity at the water inlet, is the pumping power of the water pump, is the density of river water, is the water inlet diameter of the pump.

[0059] Step S103, using the sediment disturbance core parameters to screen the initial data, to obtain the pumping data related to the sediment disturbance.

[0060] In this embodiment, the parameters that have a relatively large impact on the disturbance of the bottom sediment have been identified through correlation analysis. These parameters are used to screen the initial data according to the classification of the initial data. The pumping data screened by this method can fully represent the relevant data on the disturbance of the bottom sediment when pumping water into the river.

[0061] Step S2, obtaining a sample set according to the pumping data and the disturbance degree.

[0062] Wherein, obtaining the sample set according to the pumping data and the disturbance degree specifically includes the following sub-steps: Step S201, removing abnormal values ​​and interpolating missing values ​​from the pumping data and the disturbance degree to obtain first optimized data.

[0063] By removing outliers, we can eliminate outliers caused by equipment failure, measurement errors or extreme events as much as possible, thereby improving the reliability of data and the accuracy of analysis. The outlier removal algorithm is used to remove outliers from pumping data. Outliers refer to data that deviates significantly from other pumping data in the pumping data.

[0064] In this embodiment, the K-nearest neighbor distance method is used to eliminate outliers. The principle of the K-nearest neighbor distance method is to calculate the Euclidean distance between each data point in the pumping data and other data points. The distribution of the K nearest neighbor data points around the data point is used to determine whether it is an outlier. If the distance between a data point and its K nearest neighbor data points is significantly greater than the distance between other points, then it may be an outlier.

[0065] The Euclidean distance satisfies the following formula: , In the formula, is the Euclidean distance, and For different pumping data, is the position of the data point in the pumping data series, is the number of sampled data.

[0066] By interpolating missing values, the integrity of pumping data can be ensured, analysis bias or insufficient model training caused by missing pumping data can be avoided, and the availability of data can be enhanced.

[0067] The first optimized data refers to the data after eliminating outliers and interpolating missing values ​​in the pumping data and the disturbance degree.

[0068] In this embodiment, the mean interpolation method is used to interpolate missing values. When there are missing values ​​in the pumping data, the mean value of the variable is used to fill the missing values. This method is simple and intuitive, and is suitable for situations where the data distribution is relatively uniform and the proportion of missing values ​​is relatively small.

[0069] The mean interpolation method interpolates missing values ​​to meet the following formula: , In the formula, Indicates the pending The Pumping data for each parameter, Indicates The Pumping data for each parameter, Indicates The Pumping data for each parameter, Indicates the dimension sequence number of the parameter.

[0070] Step S202: normalize the first optimized data to obtain second optimized data.

[0071] In this embodiment, the essence of the normalization process is to scale the first optimized data so that it falls into a specific interval. From the characteristics of the first optimized data itself, normalization can eliminate the differences in dimensions and numerical ranges between different features. For example, the pumping flow rate is a large numerical range, while the disturbance degree is a relatively small numerical range. If normalization is not performed, during the data analysis and modeling process, features with large numerical ranges such as flow rate will have a greater impact on the model, covering up the effects of other features. Normalization gives each feature a relatively equal weight in the model, improving the accuracy and stability of the model.

[0072] In terms of model training, normalization helps speed up the convergence of the model. Many machine learning algorithms, such as the gradient descent method, can find the optimal solution more efficiently when trained on normalized data, reduce the number of iterations, and save computing resources and time costs. In addition, normalized data is more universal and comparable, which facilitates comparative analysis between different data sets and performance evaluation between different models, providing a more reliable data foundation for subsequent data analysis and decision making.

[0073] Step S203: construct an adversarial network model, and generate enhanced data according to the second optimized data and the adversarial network model.

[0074] In this embodiment, the adversarial network model consists of two parts: a generator and a discriminator. The task of the generator is to learn the distribution of real data and generate simulated data as realistic as possible by inputting random noise or latent vectors. The discriminator is responsible for determining whether the input data is real or generated by the generator.

[0075] The second optimized data is used as a sample of real data and input into the adversarial network model. During the training process, the generator learns the characteristics and distribution pattern of the second optimized data. At the same time, the generator and the discriminator engage in adversarial games. The generator strives to generate more realistic data to deceive the discriminator, while the discriminator continuously improves its discrimination ability and accurately distinguishes between real data and generated data. In this process of mutual confrontation and optimization, the two jointly improve performance.

[0076] Finally, the generator generates new data samples based on the learned patterns. These newly generated data are similar to the original second optimized data in characteristics and distribution, but not exactly the same, thus increasing the diversity of the data.

[0077] The generated augmented data enriches the original data set and solves the problem of insufficient data. In practical applications, especially when training machine learning and deep learning models, sufficient data is crucial for the model to learn accurate patterns and rules. Augmented data can expose the model to more samples, improve the model's generalization ability, and enable it to make more accurate predictions and judgments when faced with new and unseen data.

[0078] From the perspective of model performance, the introduction of the adversarial network model prompts the generator to generate high-quality simulated data, which can be used as additional training samples to help the model better learn the complex characteristics and potential relationships of the data. At the same time, the adversarial training process also enhances the robustness of the model, enabling it to cope with noise and uncertainty in the data, and improves the stability and reliability of the model.

[0079] In addition, the process of generating enhanced data also provides a new perspective for in-depth exploration of the intrinsic structure and changing patterns of the data, helping researchers discover hidden information and patterns in the data, and providing a more comprehensive and rich basis for further data analysis, model optimization, and related decision-making.

[0080] Step S204: merge the second optimized data and the enhanced data to obtain the sample set.

[0081] In this embodiment, the second optimized data and the enhanced data are merged to obtain a sample set, and the consistency of the data structure must be ensured during the merging process. The second optimized data is the original data that has been normalized, and the enhanced data is the simulated data generated by the adversarial network model. When merging, the data must be integrated according to the same data feature dimension and order to ensure that each data record can accurately correspond to the corresponding feature.

[0082] Step S3, constructing a water pump inlet prediction model, and setting a loss function for the water pump inlet prediction model according to the disturbance degree.

[0083] The step of setting a loss function for the water pump inlet prediction model according to the disturbance degree specifically includes the following sub-steps: Step S301, based on the water pumping test, sampling and testing the heavy metal content at the water outlet of the water pump.

[0084] In this embodiment, when conducting a pumping test, ensure that the pumping power remains unchanged, adjust the distance from the water inlet of the water pump to the bottom mud from large to small, and use this method to gradually increase the disturbance of the riverbed mud from zero. By disturbing the bottom mud, the bottom mud gradually enters the water inlet of the water pump with the water flow according to the degree of disturbance. The heavy metal content of the water outlet passing through the water pump is tested, and the heavy metal content of the corresponding water outlet of the water pump is recorded when the water inlet of the water pump is at different positions.

[0085] Step S302, based on the sampling test results, the disturbance of the riverbed sediment is evaluated to obtain the maximum acceptable degree of sediment disturbance when the water pump is pumping water.

[0086] In this embodiment, based on the impact of heavy metals on farmland, the maximum acceptable content of heavy metals in water during pumping is determined as an indicator. In a high-standard pumping environment, it is also acceptable to use zero heavy metal content as an indicator.

[0087] The distance from the water inlet of the water pump corresponding to the indicator to the bottom mud is determined according to the indicator, and the disturbance degree is calculated by combining the distance with other data.

[0088] It should be additionally explained that anti-sediment disturbance pumping does not strictly mean that the sediment cannot be disturbed at all. First of all, the river sediment is not entirely pollutant. It is only the heavy metals contained in the river sediment that will affect farmland. The disturbance coefficient of heavy metals is different from that of ordinary sediment. Therefore, the detection of the water outlet of the water pump in the present invention only targets the heavy metal content as an indicator.

[0089] Step S303, setting a loss function for the water pump inlet prediction model according to the maximum sediment disturbance degree and the disturbance degree.

[0090] The loss function satisfies the following formula: , in, is the loss function, is the sample size, For the The degree of sediment disturbance of each sample, The maximum acceptable disturbance of sediment during pumping. is the maximum value function.

[0091] In this embodiment, in the above formula, , indicating that the disturbance degree of the sample is within an acceptable range, which means that no additional penalty will be imposed on the model at this time, reflecting the optimization treatment of samples within a reasonable disturbance range. , indicating that the disturbance degree of the sample exceeds the acceptable range, which will impose a greater penalty on the model, and the greater the degree of excess, the heavier the penalty.

[0092] The role of the loss function is to evaluate the prediction results of the pump inlet prediction model by calculating the difference between the sediment disturbance degree of the sample and the maximum acceptable sediment disturbance degree. It will penalize samples that exceed the maximum acceptable sediment disturbance degree, thereby prompting the model to minimize this situation during the training process, so as to improve the accuracy and reliability of the model prediction and ensure that the sediment disturbance degree is within an acceptable range when the pump is pumping water.

[0093] Step S4: using the sample set to train and evaluate the water pump inlet prediction model.

[0094] In this embodiment, the sample set is first divided into a training set and a validation set in a ratio of 9:1. The training set is used to train the water pump inlet prediction model, and the validation set is used to evaluate the water pump inlet prediction model.

[0095] The data in the training set are input into the water pump inlet prediction model one by one. Based on these data features, the model gradually learns the intrinsic relationship between the various parameters of the water inlet and the operating status of the water pump by continuously adjusting its own parameters and weights. Through repeated iterative training, the water pump inlet prediction model can more accurately capture the complex relationship between the parameters of the training set, so that when faced with new pumping data, it can give reliable prediction results based on the input parameters, providing strong support for the efficient and stable operation of the water pump.

[0096] After the training is completed, the evaluation phase begins. Use the validation set data to test the prediction effect of the model. By calculating a series of evaluation indicators, such as mean square error, mean absolute error, determination coefficient, etc., the degree of deviation and goodness of fit between the model prediction value and the actual value are comprehensively measured. If the evaluation result is not ideal, it is necessary to readjust the structure and parameters of the model or optimize the sample set, and train and evaluate again until the model achieves satisfactory performance and can accurately predict the operating status of the water pump inlet, providing a reliable basis for the stable operation and fault warning of the water pump.

[0097] The mean absolute error satisfies the following formula: , in, represents the mean absolute error, represents the number of samples in the validation set, Indicates the validation set The actual value of the distance from the water inlet of the pump to the bottom mud of the sample, Indicates the validation set Predicted value of the distance from the water pump inlet to the bottom mud for each sample.

[0098] The root mean square error expression satisfies the following formula: , in, represents the root mean square error, represents the number of samples in the validation set, Indicates the validation set The actual value of the distance from the water inlet of the pump to the bottom mud of the sample, Indicates the validation set Predicted value of the distance from the water pump inlet to the bottom mud for each sample.

[0099] The determination coefficient expression satisfies the following formula: , in, represents the coefficient of determination, represents the number of samples in the validation set, Indicates the validation set The actual value of the distance from the water inlet of the pump to the bottom mud of the sample, Indicates the validation set The predicted value of the distance from the pump inlet to the bottom mud for each sample, It represents the average value of the actual distance from the water inlet of the validation set to the bottom mud.

[0100] The smaller the mean absolute error and the root mean square error, the better the prediction effect of the model.

[0101] In the coefficient of determination, when When , it means that the model fits the data completely accurately. In the regression model, the predicted value is exactly the same as the true value, and all data points are on the regression line or regression plane. This is an ideal fit state, indicating that the model can perfectly explain the changes in the dependent variable. When , it means that there is an error between the model prediction result and the true value, which is normal. The closer the determination coefficient is to 1, the better the prediction effect of the model.

[0102] Step S5, using the water pump water inlet prediction model to make a prediction, and automatically adjusting the water inlet and power of the water pump according to the prediction result.

[0103] The method of using the water pump water inlet prediction model to make predictions and automatically adjusting the water inlet and power of the water pump according to the prediction results specifically includes the following sub-steps: Step S501, acquiring real-time monitoring data, and performing prediction using the water pump inlet prediction model based on the real-time monitoring data; In this embodiment, the trained and evaluated water pump inlet prediction model is deployed in the pumping system to perform pumping operations. In order to realize automatic data collection, automatic detection equipment is deployed in the pumping system to automatically collect data such as river flow, pumping power, and the distance between the water inlet and the bottom mud, and the collected data is automatically sent to the large pumping system.

[0104] The density and particle diameter of riverbed mud are relatively resistant to change. Although existing technologies can be used to monitor the density and particle diameter of riverbed mud in real time, the equipment is expensive and not cost-effective. Therefore, a timed sampling method is adopted to collect the two data: the density and particle diameter of riverbed mud.

[0105] The acquired real-time monitoring data is normalized, and the water pump inlet prediction model uses the normalized real-time monitoring data to predict the distance between the water pump inlet and the bottom mud. The reason for normalizing the monitoring data is that in order to better train the model, the samples of the training model itself have been normalized. Therefore, the water pump inlet prediction model uses the normalized real-time monitoring data to predict the distance between the water pump inlet and the bottom mud, thereby improving the prediction accuracy.

[0106] Step S502, obtaining the real-time distance from the water inlet to the water surface, and automatically adjusting the water inlet and power of the water pump according to the real-time distance and the predicted result.

[0107] The automatic adjustment of the water inlet and power of the water pump according to the distance and the predicted result specifically includes the following sub-steps: Step S50201, according to the working condition of the water pump, setting the minimum distance from the water inlet of the water pump to the water surface when the water pump is working normally as a threshold; In this embodiment, setting the minimum distance from the water inlet to the water surface when the water pump is working normally as a threshold is a key measure to ensure stable, efficient and reasonable operation of the water pump.

[0108] On the one hand, setting this threshold is intended to ensure that the water pump is always in good working condition. The normal operation of the water pump depends on sufficient and stable water inlet. If the distance from the water inlet to the water surface is too large, the water pump may inhale air, causing cavitation, damaging the impeller and other key components of the water pump, and reducing the working efficiency and service life of the water pump. At the same time, it may also cause the water output of the water pump to be unstable, affecting the normal operation of the entire pumping system. By setting the minimum distance threshold, these problems can be effectively avoided and the water pump can be guaranteed to work continuously and reliably.

[0109] On the other hand, because the water inlet of the pump is adjusted according to the prediction results of the water inlet prediction model of the pump, when there is no rain for a long time or the rainfall is small, the water level of the river drops sharply. Without setting constraints, the water inlet of the pump may be forced to jump out of the water according to the prediction results.

[0110] Step S50202: automatically adjust the water inlet and power of the water pump according to the distance and the threshold.

[0111] In this embodiment, in order to ensure the irrigation of farmland, the pumping power of the water pump is generally adjusted to the maximum value under normal operation. However, when the water level gradually drops, the distance between the water inlet of the water pump and the bottom mud will gradually decrease. When the distance reaches the maximum value of the acceptable bottom mud disturbance, the water inlet will not be able to continue to drop. Because of the threshold limit, the pumping system can only adjust the power of the water pump. By reducing the power of the water pump to meet the threshold limit, it will not increase the further impact on the bottom mud disturbance, thereby realizing automatic adjustment of the water inlet and power of the water pump.

[0112] like Figure 2 As shown, on the other hand, the present invention also provides a water pump automatic adjustment system for preventing muddy bottom disturbance in a river channel, comprising: a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, and the computer program includes program instructions, and the processor is configured to call the program instructions to execute relevant steps of a relevant embodiment of a water pump automatic adjustment method for preventing muddy bottom disturbance in a river channel according to the present invention.

[0113] In the water pump automatic regulating system for preventing muddy bottom disturbance in a river provided by the present invention, each functional component can be integrated into one processing component, or each component can exist physically separately, or two or more components can be integrated into one component. The above-mentioned integrated components can be implemented in the form of hardware or in the form of software functions, further improving the overall applicability and practical application ability of the present invention.

[0114] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for automatically adjusting water pumping to prevent disturbance of muddy bottom in a river channel, characterized in that: The method comprises: Acquiring pumping data related to mud bottom disturbance, and evaluating the impact of pumping on riverbed mud based on the pumping data to obtain the degree of disturbance; Obtaining a sample set according to the pumping data and the disturbance degree; Constructing a water pump inlet prediction model, and setting a loss function for the water pump inlet prediction model according to the disturbance degree; Using the sample set, training and evaluating the water pump inlet prediction model; The water pump water inlet prediction model is used to perform prediction, and the water inlet and power of the water pump are automatically adjusted according to the prediction result.

2. A method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river according to claim 1, characterized in that: The acquisition of pumping data related to mud bottom disturbance includes: Conducting pumping tests and testing the sediments of the riverbed to obtain initial data; Performing correlation analysis on the initial data to obtain core parameters of sediment disturbance; The initial data is screened using the sediment disturbance core parameters to obtain the pumping data related to the sediment disturbance.

3. A method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river according to claim 2, characterized in that: The correlation analysis of the initial data to obtain the core parameters of sediment disturbance includes: According to the initial data, obtaining sediment disturbance parameters; Calculating the correlation coefficients between the sediment disturbance parameters, and constructing a theoretical correlation coefficient matrix based on the correlation coefficients; Performing eigenvalue decomposition on the theoretical correlation coefficient matrix to obtain eigenvectors and eigenvalues; Calculating the contribution of the sediment disturbance parameter to the eigenvector according to the eigenvector and the eigenvalue; The sediment disturbance parameters are screened using the contribution to obtain the sediment disturbance core parameters.

4. The method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river according to claim 1, characterized in that: The obtaining of a sample set according to the pumping data and the disturbance degree comprises: Eliminate outliers and interpolate missing values ​​from the pumping data and the disturbance degree to obtain first optimized data; Normalizing the first optimized data to obtain second optimized data; Constructing an adversarial network model, and generating enhanced data according to the second optimized data and the adversarial network model; The second optimized data and the enhanced data are combined to obtain the sample set.

5. The method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river according to claim 1, characterized in that: The setting of a loss function for the water pump inlet prediction model according to the disturbance degree comprises: Based on the water pumping test, sampling and testing the heavy metal content at the water outlet of the water pump; According to the sampling test results, the disturbance of the riverbed sediment is evaluated to obtain the maximum acceptable degree of sediment disturbance when the water pump is pumping water; A loss function is set for the water pump inlet prediction model according to the maximum sediment disturbance degree and the disturbance degree.

6. The method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river according to claim 1, characterized in that: The method of using the water pump water inlet prediction model to make predictions and automatically adjusting the water inlet and power of the water pump according to the prediction results includes: Acquire real-time monitoring data, and make predictions using the water pump inlet prediction model based on the real-time monitoring data; The real-time distance from the water inlet to the water surface is obtained, and the water inlet and power of the water pump are automatically adjusted according to the real-time distance and the predicted result.

7. A method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river according to claim 6, characterized in that: The automatic adjustment of the water inlet and power of the water pump according to the real-time distance and the predicted result includes: According to the working condition of the water pump, the minimum distance from the water inlet of the water pump to the water surface when the water pump works normally is set as a threshold; The water inlet and power of the water pump are automatically adjusted according to the real-time distance and the threshold.

8. The method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river channel according to claim 1, characterized in that: The disturbance degree satisfies the following formula: ， in, is the degree of mud bottom disturbance, is the comprehensive correction factor, is the pumping power of the water pump, is the distance from the water pump inlet to the mud bottom, is the river flow velocity, is the acceleration due to gravity, is the average diameter of mud bottom particles, is the density of the mud bottom, is the density of river water, is the water flow velocity at the water inlet.

9. The method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river according to claim 5, characterized in that: The loss function satisfies the following formula: ， in, is the loss function, is the sample size, For the The degree of sediment disturbance of each sample, The maximum acceptable level of sediment disturbance when pumping water.

10. A water pump automatic regulating system for preventing disturbance of river mud bottom, characterized in that: include: A processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute a method for automatically adjusting water pumping to prevent muddy bottom disturbance in a river channel as described in any one of claims 1 to 9.

Images