Multi-sensor integrated reservoir area water level monitoring method, system, equipment and medium

Abstract

The invention relates to the technical field of reservoir water level monitoring, and discloses a multi-sensor integrated reservoir water level monitoring method, system and device and a medium, and the method comprises the steps: obtaining the measurement values of all water level monitoring sensors in a target reservoir range, and carrying out the first elimination operation of the measurement values of all water level monitoring sensors; establishing a measurement input vector according to the measurement values of the plurality of water level monitoring sensors after the first elimination operation is completed; and establishing a filter set, and inputting the measurement input vector into the filter set to obtain a target state value. The precision of water level monitoring is improved, and the stability and reliability of the system are remarkably enhanced. In practical application, the system can accurately reflect the change condition of the water level of the reservoir area in real time, and provides an important decision basis for reservoir management, flood control scheduling and the like.

Description

Technical Field

[0001] This invention relates to the field of reservoir water level monitoring technology, and in particular to a multi-sensor integrated method, system, equipment and medium for reservoir water level monitoring. Background Technology

[0002] Water level monitoring is a crucial aspect of reservoir dam safety monitoring. Traditional methods rely on physical water gauges and manual readings, which are inefficient. With the continuous development of technologies such as computers, network communication, and remote sensing, water level monitoring in reservoir areas can now be mainly categorized into three types: sensor-based water level monitoring, video surveillance-based water level monitoring, and remote sensing image processing-based water level monitoring.

[0003] Among these methods, video-based water level monitoring uses cameras to identify water level gauge markings. Its advantages include low cost, but its disadvantages include complex calculations and accuracy significantly affected by lighting conditions and image quality. Water level monitoring based on remote sensing image processing compares and analyzes satellite remote sensing images from different periods to determine reservoir water levels. Its advantages include low cost and the ability to calculate reservoir area, but its disadvantages include complex calculations and low accuracy.

[0004] Sensor-based water level monitoring uses contact and non-contact water level sensors, such as float-type water level gauges, pressure-type water level gauges, ultrasonic water level gauges, radar water level gauges, and laser water level gauges, to measure water levels. Its advantages are high real-time performance, high accuracy, and the ability to measure underwater height and monitor underwater sedimentation. Its disadvantage is that single-sensor monitoring is prone to failure, leading to inaccurate water level monitoring. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] Therefore, the present invention provides a multi-sensor integrated reservoir water level monitoring method, system, equipment and medium, which can solve the problem that single sensor monitoring is prone to failure, resulting in inaccurate water level monitoring, and improve the accuracy and reliability of water level monitoring.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a multi-sensor integrated method for monitoring reservoir water levels, comprising:

[0009] Obtain the measurement values ​​of all water level monitoring sensors within the target reservoir area, and perform a first rejection operation on the measurement values ​​of all water level monitoring sensors;

[0010] A measurement input vector is established based on the measurement values ​​of several water level monitoring sensors after the first rejection operation is completed;

[0011] Establish a filter set, which includes a main filter and several sub-filters;

[0012] The measurement input vector is input into a filter set to obtain the target state value, which is the water level measurement value for this time.

[0013] The plurality of sub-filters are connected in parallel and then in series with the main filter, and the measurement input vector is the input of the plurality of sub-filters;

[0014] The outputs of the several sub-filters constitute the input of the main filter.

[0015] As a preferred embodiment of the multi-sensor integrated reservoir water level monitoring method described in this invention, the first rejection operation includes:

[0016] Obtain the precision and accuracy of all water level monitoring sensors;

[0017] The reliability information entropy of the water level monitoring sensor is calculated based on the precision and accuracy.

[0018] The reliability information entropy is sorted, and the sensor data with the lowest reliability information entropy is removed.

[0019] As a preferred embodiment of the multi-sensor integrated reservoir water level monitoring method of the present invention, the first rejection operation further includes:

[0020] The output of the remaining sub-filters after removing the sensor data with the lowest reliability entropy is subjected to outlier removal, which is performed using the box plot method.

[0021] The outlier removal operation is a data removal operation within the sensor.

[0022] As a preferred embodiment of the multi-sensor integrated reservoir water level monitoring method described in this invention, the establishment of the filter set includes:

[0023] The number of sub-filters in the filter set is determined by the number of sensors remaining after removing the sensor data with the lowest reliability information entropy;

[0024] The main filter in the filter set is connected in series with the output of the full-value filter.

[0025] This preferred solution fully utilizes data from remaining sensors, improving the accuracy and reliability of water level monitoring. By dynamically adjusting the number of sub-filters based on the number of remaining sensors, it ensures that data from each sensor is effectively used, avoiding data waste. Simultaneously, the main filter, connected in series after the outputs of all sub-filters, allows for further filtering of all sub-filter outputs, thereby further improving the accuracy and stability of water level monitoring data. This preferred solution not only improves monitoring efficiency but also enhances the system's robustness and adaptability.

[0026] As a preferred embodiment of the multi-sensor integrated reservoir water level monitoring method described in this invention, the step of acquiring the precision and accuracy of all water level monitoring sensors includes:

[0027] Obtain the average value of different parameters in each water level monitoring sensor and the observed value for the corresponding parameter;

[0028] Determine the maximum and minimum possible values ​​of the corresponding parameters based on historical experience;

[0029] The normalized variance is obtained by taking the average value of different parameters in each water level monitoring sensor, the observed value of the corresponding parameter, the maximum possible value and the minimum possible value, and combining the total number of observations.

[0030] Based on the normalized variance and the number of measurements, the average variance is determined.

[0031] The precision of the water level monitoring sensor is obtained based on the average variance and the scaling factor.

[0032] As a preferred embodiment of the multi-sensor integrated reservoir water level monitoring method of the present invention, wherein: the step of inputting the measurement input vector into a filter set to obtain the target state value includes:

[0033] Determine the filter gain of the filter, which includes the filter gain of the main filter and the filter gain of the sub-filter;

[0034] The state prediction vector and prediction mean square error are calculated based on the measured input vector, and the state estimate and mean square error are calculated in combination with the filter gain.

[0035] As a preferred embodiment of the multi-sensor integrated reservoir water level monitoring method of the present invention, the step of calculating the reliability information entropy of the water level monitoring sensor based on the precision and accuracy includes:

[0036] Based on the precision and accuracy, and combined with the total number of water level monitoring sensors, the reliability information entropy of the water level monitoring sensors is calculated.

[0037] Secondly, the present invention provides a multi-sensor integrated reservoir water level monitoring system, comprising:

[0038] The data acquisition module is used to acquire the measurement values ​​of all water level monitoring sensors within the target reservoir area, and to perform a first rejection operation on the measurement values ​​of all water level monitoring sensors.

[0039] The rejection module is used to establish a measurement input vector based on the measurement values ​​of several water level monitoring sensors after the first rejection operation is completed;

[0040] A filter creation module is used to create a filter set, which includes a main filter and several sub-filters;

[0041] The state value acquisition module is used to input the measurement input vector into a filter set to obtain the target state value, which is the water level measurement value for this time.

[0042] The plurality of sub-filters are connected in parallel and then in series with the main filter, and the measurement input vector is the input of the plurality of sub-filters;

[0043] The outputs of the several sub-filters constitute the input of the main filter.

[0044] Thirdly, the present invention provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described above.

[0045] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention proposes a multi-sensor integrated reservoir water level monitoring method, which acquires the measurement values ​​of all water level monitoring sensors within the target reservoir area, and performs a first rejection operation on the measurement values ​​of all water level monitoring sensors; establishes a measurement input vector based on the measurement values ​​of several water level monitoring sensors after the first rejection operation; establishes a filter set, inputs the measurement input vector into the filter set, and obtains the target state value.

[0047] First, by comprehensively collecting measurements from all water level monitoring sensors within the target reservoir area, this invention ensures the comprehensiveness and accuracy of the data, providing a solid foundation for subsequent processing. Second, the first rejection operation effectively removes abnormal or unreliable measurements, further improving data reliability and avoiding the impact of erroneous data on the final result. Next, a measurement input vector is established based on the rejected valid measurements, providing precise input information for filtering. By establishing a filter set including a main filter and several sub-filters, this invention can process the input vector more finely, effectively extracting and fusing information from different water level monitoring sensors, thereby obtaining a more accurate target state value, i.e., the current water level measurement value. Each step is closely linked, collectively constituting the efficient and accurate process of the multi-sensor integrated reservoir water level monitoring method of this invention.

[0048] In summary, this invention not only improves the accuracy of water level monitoring but also significantly enhances the stability and reliability of the system. In practical applications, the system can accurately reflect changes in reservoir water levels in real time, providing crucial decision-making support for reservoir management and flood control scheduling. Furthermore, the system structure of this invention is clear, easy to maintain and upgrade, and possesses good scalability and adaptability, capable of meeting the water level monitoring needs of reservoirs of different sizes and types. Therefore, the multi-sensor integrated reservoir water level monitoring system of this invention has broad application prospects and significant practical value in the field of water resource management. Attached Figure Description

[0049] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a flowchart illustrating a multi-sensor integrated method for monitoring reservoir water levels, as provided in one embodiment of the present invention.

[0051] Figure 2 This is a data flow diagram illustrating the process of eliminating failed sensors in a multi-sensor integrated reservoir water level monitoring method, as provided in one embodiment of the present invention.

[0052] Figure 3 The present invention provides a measurement input vector flow for a multi-sensor integrated reservoir water level monitoring method according to one embodiment of the present invention.

[0053] Figure 4 The filtering process of the sub-filter and main filter of a multi-sensor integrated reservoir water level monitoring method provided in one embodiment of the present invention.

[0054] Figure 5 This is an internal structure diagram of an electronic device for a multi-sensor integrated reservoir water level monitoring method provided in one embodiment of the present invention. Detailed Implementation

[0055] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0056] Example 1, referring to Figures 1-4 This is the first embodiment of the present invention, which provides a multi-sensor integrated method for monitoring reservoir water levels, including:

[0057] Existing technologies have some problems, such as the susceptibility of single-sensor monitoring to failure, leading to inaccurate water level monitoring.

[0058] This invention provides a method that can effectively solve the problems mentioned above. The following will describe in detail how to implement this multi-sensor integrated reservoir water level monitoring method with reference to several embodiments.

[0059] Figure 1 A flowchart of a multi-sensor integrated reservoir water level monitoring method is shown, including:

[0060] S101, acquire the measurement values ​​of all water level monitoring sensors within the target reservoir area, and perform the first rejection operation on the measurement values ​​of all water level monitoring sensors;

[0061] It should be noted that in order to achieve reservoir water level monitoring and overcome the shortcomings of existing technologies, data analysis needs to be conducted based on the specific target reservoir area to obtain fixed patterns.

[0062] In some specific implementations, it is necessary to collect the measurement values ​​of all water level monitoring sensors within the target reservoir area.

[0063] In some specific implementations, obtaining the measurement values ​​of all water level monitoring sensors within the target reservoir area can be achieved through direct data acquisition. The specific steps are as follows:

[0064] 1. Install various types of water level monitoring sensors (such as float-type water level gauges, pressure-type water level gauges, radar water level gauges, etc.) in the target reservoir area.

[0065] 2. Each sensor automatically collects data at preset time intervals (e.g., every 5 seconds) and transmits this data to the central processing system. This process typically requires ensuring synchronization between the sensors to guarantee data consistency and accuracy.

[0066] 3. Perform a preliminary check on the data from each sensor to ensure that the data format is correct and there are no obvious errors.

[0067] In some specific implementations, obtaining the measurement values ​​of all water level monitoring sensors within the target reservoir area can also be achieved by establishing a remote monitoring and data collection system. The specific steps are as follows:

[0068] 1. Install various types of water level monitoring sensors (such as float type, pressure type, radar water level gauge, etc.) at selected locations within the reservoir area to ensure that these sensors can cover the entire monitoring area and are in optimal working condition.

[0069] 2. Configure an appropriate communication module (such as Wi-Fi, cellular network or LoRaWAN) for each sensor and set a unified data transmission protocol (such as MQTT, HTTP, etc.) to send the collected data to a central server or cloud platform.

[0070] 3. Establish a stable central control system or utilize an existing cloud service platform to receive data streams from various sensors, and set up a database management system to store historical data, supporting query and analysis functions.

[0071] 4. The sensors automatically collect data at preset time intervals and upload it to the central control system via the network. During this process, real-time monitoring of the working status of all sensors is achieved, including key indicators such as signal strength, battery level, and data upload frequency.

[0072] 5. After initial screening of the received data to remove obvious errors or outliers, statistical methods or other advanced algorithms are used to conduct in-depth analysis of the data and extract valuable information.

[0073] 6. Finally, a user-friendly interface was developed, allowing administrators to view the latest water level information and historical trend charts, and to establish an alarm mechanism. When an anomaly is detected (such as the water level exceeding the safe range), the system automatically generates an alarm and notifies relevant personnel via SMS, email, etc. This completes a full process from data collection to processing and then to early warning.

[0074] In this embodiment of the invention, a direct data acquisition method is selected.

[0075] It should be noted that after acquiring large-scale sensor data, the sensor data needs to be processed initially to reduce errors and improve the reliability and accuracy of the data.

[0076] In some specific implementations, the first removal operation can be performed using existing techniques, such as removing outliers based on statistical methods, specifically by setting a threshold or using the 3σ principle.

[0077] First, calculate the mean and standard deviation of all sensor measurements. Then, determine a reasonable range (such as the mean plus or minus two or three times the standard deviation). Measurements outside this range are considered outliers and are removed.

[0078] This process helps to remove extremely inaccurate data caused by sensor malfunctions, environmental interference, and other factors, providing a more reliable data foundation for subsequent analysis.

[0079] In addition, the first elimination operation can also combine the experience and knowledge of domain experts to manually review and adjust data under specific conditions to ensure the accuracy and rationality of the data.

[0080] In some specific implementations, the initial removal operation can also be performed using machine learning algorithms. For example, support vector machines (SVM) or random forest algorithms can be used to classify sensor data, identify, and remove outliers that significantly deviate from the majority of data points. This method can automatically adapt to different data distributions and characteristics, improving the accuracy and efficiency of the removal operation. The application of machine learning algorithms can further reduce human intervention and increase the speed and automation of data processing.

[0081] However, in this invention, the focus is on analyzing the specific data itself. Therefore, instead of using thresholds or machine learning methods, a method that comprehensively evaluates precision and accuracy is designed for the first elimination operation. Figure 2 As shown.

[0082] In this embodiment of the invention, the first rejection operation includes:

[0083] Obtain the precision and accuracy of all water level monitoring sensors;

[0084] The reliability information entropy of the water level monitoring sensor is calculated based on its precision and accuracy.

[0085] The reliability information entropy is sorted, and the sensor data with the lowest reliability information entropy is removed.

[0086] In this embodiment of the invention, obtaining the precision and accuracy of all water level monitoring sensors includes:

[0087] Obtain the average value of different parameters in each water level monitoring sensor and the observed value for the corresponding parameter;

[0088] Determine the maximum and minimum possible values ​​of the corresponding parameters based on historical experience;

[0089] The normalized variance is obtained by taking the average value of different parameters in each water level monitoring sensor, the observed value of the corresponding parameter, the maximum possible value and the minimum possible value, and combining the total number of observations.

[0090] The mean variance is determined based on the normalized variance and the number of measurements.

[0091] The precision of the water level monitoring sensor is obtained by combining the average variance with the scaling factor.

[0092] In this embodiment of the invention, calculating the reliability information entropy of the water level monitoring sensor based on precision and accuracy includes:

[0093] Based on precision and accuracy, and combined with the total number of water level monitoring sensors, the reliability information entropy of the water level monitoring sensors is calculated.

[0094] Specifically, the sensor precision P(ε) is calculated using the following formula:

[0095]

[0096] In the formula, K is the scaling factor. ε is the average variance. i To normalize the variance, p is the number of measurements, and k is the total number of observations. v is the average value of the i-th parameter. i,n v is the nth observation of the i-th parameter. i,max ,v i,min These represent the maximum and minimum possible values ​​of this parameter under empirical conditions.

[0097] Furthermore, the sensor accuracy C(X) is calculated using the following formula:

[0098]

[0099] In the formula, n is the number of partitions in the measurement dataset, and p i This represents the probability of the data distribution within the corresponding interval.

[0100] Furthermore, the sensor reliability information entropy F is calculated using the following formula:

[0101]

[0102] In the formula, k is the total number of sensors, and P i(ε) represents the precision of sensor i, C i (X) represents the accuracy of sensor i.

[0103] Furthermore, the reliability information entropy F of all sensors is sorted.

[0104] Furthermore, sensor data with the lowest reliability entropy is eliminated.

[0105] In some specific implementations, the sorting of all sensor reliability information entropy F can be done in ascending or descending order, depending on the requirements of the embodiment.

[0106] In this embodiment of the invention, the sensor data with the lower reliability information entropy is arranged in ascending order, that is, the sensor data with the lower reliability information entropy is arranged at the front. These data are often the most unreliable and should therefore be eliminated.

[0107] In some specific implementations, the removal operation can be performed automatically or manually triggered by the system administrator.

[0108] In this embodiment of the invention, the rejection operation is performed automatically. Once the reliability information entropy of all sensors is calculated and sorted, the system automatically rejects a preset number of sensor data with the lowest reliability information entropy (e.g., the top 10%). This ensures that the system can efficiently and accurately complete the rejection operation when processing large amounts of data, reducing human intervention and errors.

[0109] It should be noted that the above elimination process only removes sensors that do not meet the conditions. A second elimination process is still required on the relevant data obtained by the remaining sensors.

[0110] In some specific implementations, secondary rejection aims to remove data from each sensor that does not meet the requirements. For example, a data quality threshold can be set, and the remaining data from each sensor can be compared with this threshold. Data below the threshold is considered low-quality data and rejected. Alternatively, a Dynamic Time Warping (DTW) algorithm can be used to compare the similarity of sensor data, identifying and removing outliers that are significantly inconsistent with other sensor data. The DTW algorithm can consider the scaling and distortion of time-series data along the time axis, thus more accurately assessing the similarity between data and improving the accuracy of the rejection operation. By combining multiple rejection strategies, this invention can comprehensively and effectively remove outlier data, providing a more reliable dataset for subsequent processing.

[0111] In some specific implementations, when a set data quality threshold is used for secondary removal, the specific steps can be as follows:

[0112] First, ensure that the reliability information entropy of the sensors has been calculated and that unreliable sensor data has been eliminated based on the sorting results.

[0113] The measurement input vector is constructed based on the remaining valid sensor data.

[0114] Based on business needs or historical data analysis, determine a reasonable data quality standard or threshold. This threshold can be set based on indicators such as error range and fluctuation amplitude.

[0115] The remaining data from each sensor is compared to a set data quality threshold; data below this threshold is considered low-quality and discarded. This step can be automated using scripts to improve efficiency and accuracy.

[0116] In some specific implementations, when the secondary removal uses a dynamic time warping algorithm, the specific steps can be as follows:

[0117] Similarly, the calculation of the sensor reliability information entropy and the corresponding data removal should be completed first.

[0118] A measurement input vector is created based on the valid sensor data that has been initially screened.

[0119] For each pair of sensor time series data, the DTW algorithm is applied to calculate the similarity distance between them. DTW is a method for comparing two time series that allows for scaling and distortion along the time axis, making it well-suited for handling time series data of varying lengths or speeds.

[0120] According to the results of the DTW algorithm, if the data from some sensors are significantly inconsistent with the data from most other sensors, these data are considered to be outliers and are removed from the dataset.

[0121] In this embodiment of the invention, the first rejection operation further includes:

[0122] The output of the remaining sub-filters after removing the sensor data with the lowest reliability entropy is subjected to outlier removal. The outlier removal operation is performed using the box plot method.

[0123] Outlier removal is a data removal operation within the sensor.

[0124] Specifically, the detailed steps for outlier removal using box plots are as follows:

[0125] First, collect all sensor measurement data after the first rejection operation, and organize the data of each sensor into a numerical sequence;

[0126] Next, calculate the quartiles (lower quartile of Q1, median of Q2, upper quartile of Q3) and the interquartile range IQR = Q3 - Q1 for each sequence. Then, determine the upper and lower bounds based on the IQR. The lower bound is usually Q1 - 1.5IQR and the upper bound is Q3 + 1.5IQR. Any value below the lower bound or above the upper bound is considered an outlier.

[0127] Next, the data sequence of each sensor is traversed, and all values ​​that are less than the lower bound or greater than the upper bound are marked as potential outliers. For these data points marked as outliers, they are directly removed from the dataset or replaced with missing values ​​such as NaN.

[0128] Finally, new measurement input vectors are constructed using the processed data to ensure that the data used for further analysis is more accurate and reliable. If necessary, box plots can be applied again after the initial removal to ensure that no significant outliers are missed. This method is simple and effective, and suitable for data cleaning in the preprocessing stage.

[0129] It should be noted that acquiring the measurements from all water level monitoring sensors within the target reservoir area and performing an initial filtering operation on these measurements can improve the accuracy and reliability of the data. After this initial filtering, the remaining data in the system will more closely reflect the actual water level situation, reducing the impact of outliers or erroneous data on subsequent analysis and decision-making. This helps the system more accurately reflect the actual water level situation in the reservoir area, providing more reliable data support for water level monitoring, early warning, and scheduling. Simultaneously, the dataset after initial filtering also lays a solid foundation for further data cleaning and analysis that may be conducted later.

[0130] S102, establish a measurement input vector based on the measurement values ​​of several water level monitoring sensors after the first rejection operation is completed;

[0131] In some specific implementations, the steps for establishing a measurement input vector based on the measurements from several water level monitoring sensors after the first rejection operation can be as follows:

[0132] 1. First, ensure that the initial screening of all water level monitoring sensors has been completed (first elimination operation). This step includes calculating the precision and accuracy of each sensor and eliminating unreliable sensor data based on the reliability information entropy.

[0133] 2. Collect all measurement data from the remaining valid sensors after the first elimination operation. For example, suppose that after elimination, there are 3 sensors left, labeled as S1, S2, and S3, and each sensor collected 12 data points, with datasets D1, D2, and D3 respectively.

[0134] 3. For each sensor dataset (e.g., D1, D2, D3), perform outlier removal. This can be done using a box plot method. The steps include calculating the quartiles Q1, Q3, and IQR (interquartile range) for each dataset, and then determining the upper and lower bounds (typically Q1 - 1.5IQR and Q3 + 1.5IQR). Any values ​​exceeding these boundaries are considered outliers and removed.

[0135] For example, when processing the dataset D1 of sensor S1, first calculate its Q1, Q3 and IQR, and then find and remove outliers in D1.

[0136] 4. After removing outliers, calculate the average of the remaining data for each sensor as the current measurement value for that sensor. For example, for sensor S1, if there are 10 valid measurements remaining after outlier removal, take the average of these 10 values ​​as the current measurement value for S1.

[0137] 5. Combine the current measurement values ​​of all remaining sensors into a vector, which is the measurement input vector. Continuing the example above, if only two sensors S1 and S2 remain, combine the current measurement values ​​of these two sensors into a 1×2 vector, such as [avg(D1),avg(D2)], where avg(D1) and avg(D2) are the current measurement values ​​of sensors S1 and S2, respectively.

[0138] 6. Ensure that the measurement input vector contains all necessary information and does not include any known anomalies or invalid data. If necessary, adjust the selection of which sensor data is included in the final measurement input vector based on the actual situation.

[0139] It should be noted that, as Figure 3 As shown, the most reliable data can be filtered from multiple sensors to form a precise measurement input vector for subsequent processing (such as filter processing). This method not only improves data quality but also provides a solid foundation for more accurate water level monitoring.

[0140] S103, Establish a filter set, which includes a main filter and several sub-filters;

[0141] It should be noted that after obtaining the target measurement input vector, further processing is required to extract water level feature information. To this end, this invention designs a filter set, consisting of a main filter and several sub-filters. The main filter is responsible for processing the main part of the measurement input vector and extracting key water level features. The sub-filters, on the other hand, process specific data features or anomalies to further enhance the accuracy and reliability of the data. Through the collaborative work of the main filter and sub-filters, this invention can achieve comprehensive and accurate monitoring of water level data. In practical implementation, the selection and design of the main filter and sub-filters will be determined based on actual application requirements and the characteristics of water level changes in the reservoir area.

[0142] In this embodiment of the invention, establishing the filter set includes:

[0143] The number of sub-filters in the filter set is determined by the number of sensors remaining after removing the sensor data with the lowest reliability entropy;

[0144] The main filter in the filter set is connected in series with the output of the full-value filter.

[0145] Specifically, such as Figure 4 As shown, the present invention employs a two-stage hybrid federated filtering system comprising m sub-filters and one master filter. The outputs of the m sub-filters serve as the inputs to the master filter, and the output of the master filter serves as the observed value of the reservoir water level at the current moment. Both the sub-filters and the master filter employ Kalman filters.

[0146] Furthermore, the Kalman filter model formula used is as follows:

[0147]

[0148] In the formula, This represents the state vector of the i-th sub-filter in the k-th iteration. This represents the state transition matrix of the i-th sub-filter from the (k-1)-th iteration to the k-th iteration. This represents the state vector of the i-th sub-filter in the (k-1)-th iteration. Let represent the noise allocation matrix of the i-th sub-filter in the (k-1)-th iteration. This represents the system noise vector of the i-th sub-filter in the (k-1)-th iteration. The measurement vector of the i-th sub-filter in the k-th iteration. The measurement matrix of the i-th sub-filter in the k-th iteration Let represent the noise vector measured in the k-th iteration of the i sub-filters.

[0149] Preferably, the filtering process for each filter from the (k-1)th to the kth iteration includes the following steps:

[0150] Furthermore, the state prediction vector is calculated. The formula is:

[0151]

[0152] In the formula, Φ k，k-1 Let X be the state transition matrix from the (k-1)th iteration to the kth iteration. k-1 This is the state vector for the (k-1)th iteration;

[0153] Furthermore, the prediction mean square error P is calculated. k，k-1 The formula is:

[0154]

[0155] In the formula, P k-1 Let be the mean square error of the k-th iteration. Φ k,k-1 The transpose of Q k-1 W k-1 The covariance matrix of the system noise vector in the (k-1)th iteration;

[0156] Furthermore, calculate the Kalman filter gain K. k The formula is:

[0157]

[0158] In the formula, The k-th iteration measurement matrix H k The transpose of R k The measurement noise vector V for the k-th iteration k The covariance matrix;

[0159] Furthermore, the state estimate X for the k-th iteration is calculated. k The formula is:

[0160]

[0161] Furthermore, calculate the mean square error P of the k-th iteration. k The formula is:

[0162] P k =(IK k H k )P k,k-1

[0163] Preferably, the mathematical formula of the Kalman filter in the main filter is the same as that of the sub-filter.

[0164] It is important to note that establishing a filter set can improve the accuracy and reliability of water level monitoring data. By designing a filter set containing a main filter and multiple sub-filters, refined processing can be performed on different data characteristics. The main filter is responsible for processing the overall data and extracting key water level features, while the sub-filters are finely adjusted for specific anomalies or specific data features. This multi-level processing approach not only enhances data accuracy but also helps identify and eliminate potential error sources. Furthermore, using the Kalman filter as the core algorithm for both the main and sub-filters enables optimal estimation using historical data and current observations, further improving the accuracy and stability of water level monitoring. Therefore, establishing a filter set is a crucial step in achieving accurate water level monitoring, providing a reliable foundation for subsequent data analysis and decision-making.

[0165] S104, input the measurement input vector into the filter set to obtain the target state value, which is the water level measurement value for this time;

[0166] In this embodiment of the invention, several sub-filters are connected in parallel and then in series with the main filter, and the measured input vector is the input of several sub-filters;

[0167] In this embodiment of the invention, the outputs of several sub-filters constitute the input of the main filter.

[0168] In this embodiment of the invention, the measurement input vector is input into a filter set to obtain the target state value, including:

[0169] Determine the filter gain, which includes the filter gain of the main filter and the filter gain of the sub-filters;

[0170] The state prediction vector and the prediction mean square error are calculated based on the measured input vector, and the state estimate and mean square error are calculated in combination with the filter gain.

[0171] Specifically, in the prediction phase, the state at the current moment is predicted based on the state estimate from the previous moment. In the update phase, the predicted value is corrected using the actual measured value at the current moment and the Kalman gain, thus obtaining the best estimate for the current moment. Simultaneously, the mean squared error is calculated and updated, but this value is mainly used to assess the confidence level of the estimate, rather than being directly used to calculate the target state value. The final target state value is the best estimate for the current moment obtained after the prediction and update phases using the Kalman filter, reflecting the current water level situation in the reservoir area. The mean squared error is used to help understand the reliability of this estimate.

[0172] In summary, this invention proposes a multi-sensor integrated reservoir water level monitoring method, which acquires the measurement values ​​of all water level monitoring sensors within the target reservoir area and performs a first rejection operation on the measurement values ​​of all water level monitoring sensors; establishes a measurement input vector based on the measurement values ​​of several water level monitoring sensors after the first rejection operation; establishes a filter set, inputs the measurement input vector into the filter set, and obtains the target state value.

[0173] First, by comprehensively collecting measurements from all water level monitoring sensors within the target reservoir area, this invention ensures the comprehensiveness and accuracy of the data, providing a solid foundation for subsequent processing. Second, the first rejection operation effectively removes abnormal or unreliable measurements, further improving data reliability and avoiding the impact of erroneous data on the final result. Next, a measurement input vector is established based on the rejected valid measurements, providing precise input information for filtering. By establishing a filter set including a main filter and several sub-filters, this invention can process the input vector more finely, effectively extracting and fusing information from different water level monitoring sensors, thereby obtaining a more accurate target state value, i.e., the current water level measurement value. Each step is closely linked, collectively constituting the efficient and accurate process of the multi-sensor integrated reservoir water level monitoring method of this invention.

[0174] In summary, this invention not only improves the accuracy of water level monitoring but also significantly enhances the stability and reliability of the system. In practical applications, the system can accurately reflect changes in reservoir water levels in real time, providing crucial decision-making support for reservoir management and flood control scheduling. Furthermore, the system structure of this invention is clear, easy to maintain and upgrade, and possesses good scalability and adaptability, capable of meeting the water level monitoring needs of reservoirs of different sizes and types. Therefore, the multi-sensor integrated reservoir water level monitoring system of this invention has broad application prospects and significant practical value in the field of water resource management.

[0175] Example 2, in a preferred embodiment, uses three sensors—a float-type water level gauge, a pressure-type water level gauge, and a radar water level gauge—to acquire water level data at a certain collection point in a reservoir. The sensor resolution is 1 cm, and the measurement accuracy is ±0.5 cm. Each sensor collects data every 5 seconds, for a total of 12 data points, denoted as [d1d2……d12]. The three sensor measurement datasets are denoted as D1, D2, and D3, respectively.

[0176] Specifically, when calculating sensor density, K=1 indicates no scaling, p=12 indicates 12 measurements, and k=1 indicates one observation. v i,max ,v i,min The highest and lowest observed water levels at the data collection points of the reservoir were recorded respectively.

[0177] Furthermore, the measurement dataset is divided into 5 regions, i.e., n=5, to count the number of times the data set [d1d2……d12] falls in each of the 5 regions. The ratio of the number of occurrences to the total number of measurement data (12) is used as the probability p. i .

[0178] Furthermore, the sensor data with the lowest reliability entropy was removed, and only the measurement data from the two sensors were retained, ultimately resulting in a 1×2 measurement input vector.

[0179] Furthermore, the two-stage hybrid federated filtering of this invention includes two sub-filters and one master filter. The outputs of the two sub-filters serve as the inputs of the master filter, and the output of the master filter serves as the observed value of the reservoir water level at the current moment. Both the sub-filters and the master filter employ Kalman filters.

[0180] Furthermore, take Φ k,k-1 =1, initial state X0=1.

[0181] Furthermore, take Q k-1 =0.1, initial mean square error P0=10.

[0182] Furthermore, take H k =1,R k =1.

[0183] Furthermore, we take Z0 as the initial measurement value.

[0184] Furthermore, the filtering steps of the main filter are similar to those of the sub-filters. Since the input of the main filter is a 1×2 dimensional vector, we take...

[0185] Furthermore, the fused state value, i.e., the state value obtained by the main filter, is saved as the current water level measurement value. In the next measurement, a more accurate value will be calculated by comparing the current measurement value with the saved value.

[0186] Example 3, referring to Figure 5 This embodiment also provides a multi-sensor integrated reservoir water level monitoring system, including:

[0187] The data acquisition module is used to acquire the measurement values ​​of all water level monitoring sensors within the target reservoir area and to perform a first rejection operation on the measurement values ​​of all water level monitoring sensors.

[0188] The rejection module is used to establish a measurement input vector based on the measurement values ​​of several water level monitoring sensors after the first rejection operation is completed;

[0189] The filter creation module is used to create a filter set, which includes a main filter and several sub-filters.

[0190] The state value acquisition module is used to input the measurement input vector into the filter set to obtain the target state value, which is the water level measurement value for this time.

[0191] Several sub-filters are connected in parallel and then in series with the main filter. The measured input vector is the input of several sub-filters.

[0192] The outputs of several sub-filters form the input of the main filter.

[0193] The above-mentioned unit modules can be embedded in the processor of the electronic device in hardware form or independent of it, or they can be stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of the above modules.

[0194] This embodiment also provides an electronic device, which can be a terminal, and its internal structure diagram can be as follows: Figure 5 As shown, the electronic device includes a processor, memory, communication interface, display screen, and input device connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a multi-sensor integrated reservoir water level monitoring method. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the device's casing, or an external keyboard, touchpad, or mouse.

[0195] This embodiment also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, it performs the following steps:

[0196] Obtain the measurement values ​​of all water level monitoring sensors within the target reservoir area, and perform a first rejection operation on the measurement values ​​of all water level monitoring sensors;

[0197] A measurement input vector is established based on the measurement values ​​of several water level monitoring sensors after the first rejection operation is completed;

[0198] Establish a filter set, which includes a main filter and several sub-filters;

[0199] The measurement input vector is input into a filter set to obtain the target state value, which is the water level measurement value for this time.

[0200] Several sub-filters are connected in parallel and then in series with the main filter. The measured input vector is the input of several sub-filters.

[0201] The outputs of several sub-filters form the input of the main filter.

[0202] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0203] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0204] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A multi-sensor integrated method for monitoring reservoir water levels, characterized in that, include: Obtain the measurement values ​​of all water level monitoring sensors within the target reservoir area, and perform a first rejection operation on the measurement values ​​of all water level monitoring sensors; A measurement input vector is established based on the measurement values ​​of several water level monitoring sensors after the first rejection operation is completed; Establish a filter set, which includes a main filter and several sub-filters; The measurement input vector is input into a filter set to obtain the target state value, which is the water level measurement value for this time. The plurality of sub-filters are connected in parallel and then in series with the main filter, and the measurement input vector is the input of the plurality of sub-filters; The outputs of the several sub-filters constitute the input of the main filter.

2. The multi-sensor integrated reservoir water level monitoring method as described in claim 1, characterized in that, The first rejection operation includes: Obtain the precision and accuracy of all water level monitoring sensors; The reliability information entropy of the water level monitoring sensor is calculated based on the precision and accuracy. The reliability information entropy is sorted, and the sensor data with the lowest reliability information entropy is removed.

3. The multi-sensor integrated reservoir water level monitoring method as described in claim 2, characterized in that, The first rejection operation also includes: The output of the remaining sub-filters after removing the sensor data with the lowest reliability entropy is subjected to outlier removal, which is performed using the box plot method. The outlier removal operation is a data removal operation within the sensor.

4. The multi-sensor integrated reservoir water level monitoring method as described in claim 3, characterized in that, The established filter set includes: The number of sub-filters in the filter set is determined by the number of sensors remaining after removing the sensor data with the lowest reliability information entropy; The main filter in the filter set is connected in series with the output of the full-value filter.

5. The multi-sensor integrated reservoir water level monitoring method as described in claim 4, characterized in that, The acquisition of the precision and accuracy of all water level monitoring sensors includes: Obtain the average value of different parameters in each water level monitoring sensor and the observed value for the corresponding parameter; Determine the maximum and minimum possible values ​​of the corresponding parameters based on historical experience; The normalized variance is obtained by taking the average value of different parameters in each water level monitoring sensor, the observed value of the corresponding parameter, the maximum possible value and the minimum possible value, and combining the total number of observations. Based on the normalized variance and the number of measurements, the average variance is determined. The precision of the water level monitoring sensor is obtained based on the average variance and the scaling factor.

6. The multi-sensor integrated reservoir water level monitoring method as described in claim 5, characterized in that, The step of inputting the measurement input vector into a filter set to obtain the target state value includes: Determine the filter gain of the filter, which includes the filter gain of the main filter and the filter gain of the sub-filter; The state prediction vector and prediction mean square error are calculated based on the measured input vector, and the state estimate and mean square error are calculated in combination with the filter gain.

7. The multi-sensor integrated reservoir water level monitoring method as described in claim 6, characterized in that, The calculation of the reliability information entropy of the water level monitoring sensor based on the precision and accuracy includes: Based on the precision and accuracy, and combined with the total number of water level monitoring sensors, the reliability information entropy of the water level monitoring sensors is calculated.

8. A multi-sensor integrated reservoir water level monitoring system, using the method described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to acquire the measurement values ​​of all water level monitoring sensors within the target reservoir area, and to perform a first rejection operation on the measurement values ​​of all water level monitoring sensors. The rejection module is used to establish a measurement input vector based on the measurement values ​​of several water level monitoring sensors after the first rejection operation is completed; A filter creation module is used to create a filter set, which includes a main filter and several sub-filters; The state value acquisition module is used to input the measurement input vector into a filter set to obtain the target state value, which is the water level measurement value for this time. The plurality of sub-filters are connected in parallel and then in series with the main filter, and the measurement input vector is the input of the plurality of sub-filters; The outputs of the several sub-filters constitute the input of the main filter.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the multi-sensor integrated reservoir water level monitoring method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the multi-sensor integrated reservoir water level monitoring method according to any one of claims 1 to 7.

Images