Drainage efficiency intelligent regulation method and system based on drainage flow monitoring data

Abstract

This invention relates to the field of drainage flow data processing technology, specifically to an intelligent control method and system for drainage efficiency based on drainage flow monitoring data. The invention comprises a sampled data sequence and a predicted data sequence; it obtains multiple data segment groups from the sampled data sequence and the predicted data sequence based on the volatility of the data in the sampled data sequence; it obtains the prediction deviation of each data segment group based on the difference distribution characteristics and relative distance between the two data sequences in each data segment group; it obtains the relative prediction deviation difference at real time based on the difference between the prediction deviation of the data segment group at the real time and the prediction deviation of other data segment groups within the historical range; it obtains the weighted prediction data for the next time step; it further obtains the predicted gate opening degree; and it controls the drainage gate. This invention optimizes the gate opening degree and improves drainage efficiency by obtaining accurate drainage flow prediction data for the next time step.

Description

Technical Field

[0001] This invention relates to the field of drainage flow data processing technology, and more specifically to a method and system for intelligent control of drainage efficiency based on drainage flow monitoring data. Background Technology

[0002] Municipal road drainage projects are an important part of urban infrastructure construction. Drainage gates are one of the key facilities used to control water flow to prevent water accumulation and flooding from invading roads or low-lying areas, thereby ensuring traffic safety and the normal operation of the city. Therefore, regulating drainage gates is very important for improving the overall efficiency of drainage projects.

[0003] In existing technologies, regulating drainage efficiency in engineering projects based on drainage flow monitoring data relies on advanced data collection and analysis techniques to monitor and adjust the operational status of the drainage system in real time, thereby improving drainage efficiency. However, predicting the next moment's drainage flow data through time series analysis of drainage flow monitoring data may result in significant deviations between predicted and actual values ​​due to uncertainties such as equipment status. This makes it difficult to accurately adjust the opening degree of drainage gates. Under certain extreme weather conditions, closing the gates may cause the water level to rise too quickly, increasing the risk of flooding and resulting in poor drainage efficiency. Under normal circumstances, keeping the gates open may lower the water level when drainage is not needed, affecting the surrounding ecosystem. Inaccurate prediction results of drainage flow data lead to inaccurate opening degrees of drainage gates, resulting in poor drainage efficiency. Summary of the Invention

[0004] To address the technical problem of inaccurate drainage flow predictions leading to inaccurate gate opening degrees, this invention aims to provide an intelligent drainage efficiency control method and system based on drainage flow monitoring data. The specific technical solution adopted is as follows:

[0005] This invention proposes an intelligent control method for drainage efficiency based on drainage flow monitoring data, the method comprising:

[0006] The drainage flow sampling data sequence of the drainage gate at each sampling time is obtained in real time according to the preset sampling time period;

[0007] Based on the variation characteristics of the drainage flow sampling data in the sampling data sequence, the predicted drainage flow data at the next time moment for each sampling moment is obtained, forming a predicted data sequence; based on the volatility of the data in the sampling data sequence, multiple data segment groups between the sampling data sequence and the predicted data sequence are obtained; based on the difference distribution characteristics and relative distance between the two data sequences in each data segment group, the prediction deviation of each data segment group is obtained.

[0008] Based on the difference between the prediction deviation of the data segment group at the real time and the prediction deviation of other data segment groups in the historical range, the relative prediction deviation at the real time is obtained; based on the relative prediction deviation at the real time and the prediction deviation of all data segment groups, the drainage flow prediction data for the next time of the real time is adjusted to obtain the weighted prediction data for the next time of the real time.

[0009] Obtain the real-time gate opening degree at real time, and obtain the predicted gate opening degree based on the real-time drainage flow sampling data, the real-time gate opening degree, and the weighted prediction data;

[0010] The drainage gate is adjusted according to the predicted gate opening degree.

[0011] Furthermore, the method for obtaining the drainage flow prediction data includes:

[0012] For the earliest sampling time in the sampling data sequence, the drainage flow sampling data of each sampling time is used as the drainage flow prediction data of the corresponding sampling time;

[0013] Starting from the earliest sampling time in the time series, calculate the difference between the drainage flow sampling data at each sampling time and the corresponding previous sampling time, and use it as the first difference;

[0014] The sum of the first difference and the drainage flow sampling data at each sampling time is calculated as the drainage flow prediction data for the next time step at each sampling time step.

[0015] Furthermore, the method for obtaining the data segment group includes:

[0016] Based on the volatility of the data in the sampled data sequence, the APCA method is applied to the sampled data sequence to obtain multiple data segments; the data with the smallest temporal sequence in the predicted data sequence and the sampled data sequence are aligned and arranged, and the predicted data sequence is segmented according to the segmentation position of the data segments in the sampled value sequence to obtain multiple data segments of the predicted data sequence.

[0017] The corresponding data segments between the sampled data sequence and the predicted data sequence are grouped into data segment groups.

[0018] Furthermore, the method for obtaining the prediction deviation includes:

[0019] In the data segment group, the DTW algorithm is used to calculate the relative distance between the sampled data sequence and the predicted data sequence, which is used as the matching distance;

[0020] In the data segment group, the DTW algorithm is used to calculate the relative distance between the sampled data sequence and the predicted data sequence, which is used as the matching distance;

[0021] Calculate the difference between each corresponding drainage flow sampling data between the sampled data sequence and the predicted data sequence, as the data difference; calculate the mean of all data differences between the sampled data sequence and the predicted data sequence, as the first deviation coefficient;

[0022] Calculate the mean and standard deviation of all drainage flow sampling data in the sampling data sequence, and multiply them to obtain the first product; calculate the mean and standard deviation of all drainage flow sampling data in the prediction data sequence, and multiply them to obtain the second product; calculate the difference between the first product and the second product as the second deviation coefficient.

[0023] The prediction deviation of the data segment group is obtained based on the matching distance, the first deviation coefficient, and the second deviation coefficient; the matching distance, the first deviation coefficient, and the second deviation coefficient are all positively correlated with the prediction deviation.

[0024] Furthermore, the method for obtaining the relative prediction deviation difference includes:

[0025] The mean absolute value of the prediction deviation of all other data segment groups outside the data segment group at the real time is calculated as the average deviation value;

[0026] Calculate the difference between the absolute value of the prediction deviation and the average deviation value of the data segment group at the real time, and perform normalization mapping to obtain the relative prediction deviation at the real time.

[0027] Furthermore, the method for obtaining the weighted prediction data includes:

[0028] Based on the numerical characteristics of the relative prediction deviation at real time, it is determined whether the drainage flow prediction data for the next time moment needs to be adjusted. If adjustment is required, a weighting coefficient is obtained based on the relative prediction deviation at real time moment and the prediction deviation of all data segment groups.

[0029] The product of the weighting coefficient and the predicted drainage flow rate at the next time step is calculated to obtain the weighted prediction data for the next time step at the current time step.

[0030] Furthermore, determining whether the drainage flow prediction data for the next time moment needs adjustment based on the numerical characteristics of the relative prediction deviation at real-time includes:

[0031] If the relative prediction deviation at a real-time moment is not zero, the predicted drainage flow rate for the next moment needs to be adjusted; otherwise, no adjustment is required.

[0032] Furthermore, the method for obtaining the weighting coefficients is as follows:

[0033] For the relative prediction deviation at real time that is negative, the sum of the prediction deviation of the data segment group at real time and the preset constant is calculated and used as a weighting coefficient.

[0034] For the relative prediction deviation at real time, which is positive, the one with the smallest absolute value of prediction deviation among all data segment groups is selected as the minimum absolute value; the sum of the minimum absolute value and the preset constant is calculated as the weighting coefficient.

[0035] Furthermore, the method for obtaining the degree of gate opening includes:

[0036] Obtain the maximum carrying capacity of the drainage gate; if the weighted predicted data at the next moment is less than the maximum carrying capacity, calculate the ratio of the weighted predicted data at the next moment to the drainage flow sampling data at the real time, and use it as the first ratio; calculate the product of the first ratio and the real-time gate opening degree at the real time, and use it as the predicted gate opening degree at the next moment.

[0037] If the weighted predicted data at the next moment is greater than or equal to the maximum carrying capacity flow value, the maximum gate opening degree of the drainage gate will be used as the predicted gate opening degree at the next moment.

[0038] The present invention also proposes an intelligent control system for drainage efficiency based on drainage flow monitoring data, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the steps of the intelligent control method for drainage efficiency based on drainage flow monitoring data described above.

[0039] The present invention has the following beneficial effects:

[0040] This invention obtains predicted drainage flow data for each sampling moment at the next moment based on the variation characteristics of drainage flow sampling data in the sampling data sequence, forming a predicted data sequence. This captures the trend and periodic changes of drainage flow data, more accurately reflecting the actual situation of drainage flow. Based on the volatility of the data in the sampling data sequence, multiple data segment groups are obtained between the sampling data sequence and the predicted data sequence. Based on the difference distribution characteristics and relative distance between the two data sequences in each data segment group, the prediction deviation of each data segment group is obtained, quantifying the deviation between the predicted data and the sampling data, and identifying data segment groups with large prediction errors. Based on the prediction deviation of the data segment group at the real time and other data within the historical range... The invention obtains the relative prediction deviation difference at real-time by analyzing the differences in prediction deviation between data segments, accurately identifying the degree of prediction deviation at real-time. Based on the relative prediction deviation difference at real-time and the prediction deviation of all data segments, the drainage flow prediction data for the next time step is adjusted to obtain weighted prediction data for the next time step. This targeted adjustment of the drainage flow prediction data reduces prediction errors and improves prediction accuracy. The invention also obtains the real-time gate opening degree and, based on the real-time drainage flow sampling data, the real-time gate opening degree, and the weighted prediction data, obtains the predicted gate opening degree, enabling proactive control of drainage efficiency and improving the operating efficiency of the drainage system. In essence, this invention improves drainage efficiency by obtaining accurate drainage flow prediction data for the next time step and optimizing gate opening degree. Attached Figure Description

[0041] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. 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.

[0042] Figure 1 A flowchart illustrating an intelligent control method for drainage efficiency based on drainage flow monitoring data, provided in one embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of a drainage flow sampling data sequence provided in one embodiment of the present invention;

[0044] Figure 3 This is a schematic diagram of a data segment group division provided in an embodiment of the present invention. Detailed Implementation

[0045] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a drainage efficiency intelligent control method and system based on drainage flow monitoring data proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0047] The following description, in conjunction with the accompanying drawings, details the specific scheme of the intelligent control method and system for drainage efficiency based on drainage flow monitoring data provided by the present invention.

[0048] Please see Figure 1 The document illustrates a flowchart of an intelligent control method for drainage efficiency based on drainage flow monitoring data, according to an embodiment of the present invention. The specific method includes:

[0049] Step S1: Obtain the drainage flow sampling data sequence of the drainage gate at each sampling time in real time according to the preset sampling time period.

[0050] In embodiments of the present invention, in order to analyze drainage flow sampling data and monitor and adjust the operating status of the drainage system in real time to improve drainage efficiency, a flow meter or water level sensor is first installed at the inlet of the drainage gate to acquire and monitor the drainage flow in real time. The drainage flow sampling data sequence of the drainage gate at each sampling time is acquired in real time according to a preset sampling time period, such as... Figure 2 The diagram illustrates a sequence of drainage flow sampling data.

[0051] It should be noted that, in one embodiment of the present invention, the preset sampling time period is 15 minutes and the time interval is 10 seconds; in other embodiments of the present invention, the size of the preset sampling time period and the time interval can also be set according to the specific situation, and will not be limited or described in detail here.

[0052] Step S2: Based on the variation characteristics of the drainage flow sampling data in the sampling data sequence, obtain the drainage flow prediction data at different times to form the prediction data sequence corresponding to the next time moment; based on the volatility of the data in the sampling data sequence, obtain multiple data segment groups between the sampling data sequence and the prediction data sequence; based on the difference distribution characteristics and relative distance between the two data sequences in each data segment group, obtain the prediction deviation of each data segment group.

[0053] Analyzing drainage flow sampling data allows us to understand the trends and periodic patterns of drainage flow, thereby predicting drainage flow data and gaining a better understanding of the drainage system's operational status. Therefore, by analyzing the changing characteristics of drainage flow sampling data in a sampling data sequence, we can obtain predicted drainage flow data for each sampling moment at the next moment, forming a predicted data sequence that provides intuitive information about predicted drainage flow.

[0054] Preferably, in one embodiment of the present invention, the method for obtaining drainage flow prediction data includes:

[0055] For the earliest sampling time in the sampling data sequence, the drainage flow sampling data of each sampling time is used as the drainage flow prediction data of the corresponding sampling time; starting from the second earliest sampling time in the time sequence, the difference between the drainage flow sampling data of each sampling time and the corresponding previous sampling time is calculated as the first difference.

[0056] The sum of the first difference and the drainage flow sampling data at each sampling time is calculated to serve as the predicted drainage flow data for the next time step at each sampling time step. In one embodiment of the present invention, the formula for the predicted drainage flow data for the next time step is expressed as:

[0057] F i ′ +1 =F i +(F i -F i-1 );

[0058] Among them, F i ′ +1 This represents the predicted drainage flow rate at the next time point i+1 after sampling time i; F i F represents the drainage flow rate sampling data at sampling time i; i-1 This represents the drainage flow sampling data at the time i-1 preceding sampling time i.

[0059] In the predicted drainage flow rate for the next time step, (F i -F i-1 The difference between the sampling time and the previous time represents the difference in drainage flow sampling data. If the difference is positive, the larger the difference, the greater the change in drainage flow between time points, and the faster the predicted drainage flow data for the next time point increases. The smaller the difference, the smaller the change in drainage flow between time points, and the smaller the predicted drainage flow data for the next time point increases. Conversely, if the difference is negative, the larger the difference, the faster the predicted drainage flow data for the next time point decreases. The smaller the difference, the slower the predicted drainage flow data for the next time point decreases.

[0060] In one embodiment of the present invention, the method for obtaining the predicted data sequence is as follows: if there exists a sampled data sequence {1,2,4,5,6}, the predicted data is calculated to be 1,2,3,6,6,7, and the resulting predicted data sequence is {2,3,6,6,7}. It should be noted that in this embodiment of the present invention, the time range of the predicted data sequence is obtained by moving one position forward from the time range of the sampled data sequence.

[0061] In another embodiment of the present invention, if there exists a sampled data sequence of {1,2,4,5,6}, and the calculated predicted data is 1,2,3,6,6,7, the predicted data sequence can also be {1,2,3,6,6,7} for subsequent analysis. That is, in this embodiment of the present invention, the two sequences are temporally corresponding from the first time point, and the last element of the predicted data sequence is the predicted data for the next time point of the last element of the sampled data sequence.

[0062] The volatility of data in a sampled data sequence reflects the differences in the data sequence at different times, which can better identify and distinguish the different characteristics of the data; by dividing the data sequence into multiple data segment groups, the variation characteristics between the drainage flow sampled data and the predicted data in each data segment group can be analyzed more specifically, and the prediction level of the drainage system can be grasped more accurately; multiple data segment groups between the sampled data sequence and the predicted data sequence can be obtained based on the volatility of the data in the sampled data sequence.

[0063] Preferably, in one embodiment of the present invention, the method for obtaining the data segment group includes:

[0064] Based on the volatility of the sampled data sequence, the APCA method is applied to obtain multiple data segments. The predicted data sequence and the sampled data sequence are aligned and arranged with the smallest temporally significant data. The arranged predicted data sequence is then segmented according to the segmentation positions of the data segments in the sampled data sequence, resulting in multiple data segments. Corresponding data segments between the sampled and predicted data sequences are grouped into data segment groups. For example... Figure 3 This diagram illustrates a data segment grouping method; the data sequences are aligned at both ends, and the area between the dashed lines represents the range of the data segment group.

[0065] It should be noted that if a sampled data sequence {1,2,4,5,6} exists, and the data segments are divided into {1,2,4} and {5,6}, then the corresponding predicted data sequence is {2,3,6,6,7}. After aligning the smallest temporally ordered data segments between the sequences, the corresponding segments are {2,3,6} and {6,7}. {1,2,4} and {2,3,6}, and {5,6} and {2,3,6} constitute two data segment groups. APCA segmentation is based on the local characteristics and variation patterns of the data, dividing the continuous data sequence into several relatively stable sub-segments, so that the data values ​​within each sub-segment can be approximated as constants or have very small variations. The specific APCA method is a well-known technique to those skilled in the art and will not be elaborated upon here.

[0066] The difference distribution characteristics reflect the deviation distribution between the drainage flow sampling data and the predicted data, helping to identify which data segments have larger and smaller prediction deviations. The relative distance measures the relative deviation between the drainage flow sampling data and the predicted data, reflecting the deviation between data sequences more comprehensively and helping to conduct targeted analysis for different data segments. Therefore, based on the difference distribution characteristics and relative distance between the two data sequences in each data segment, the prediction deviation of each data segment is obtained.

[0067] Preferably, in one embodiment of the present invention, the method for obtaining the prediction deviation includes:

[0068] In the data segment group, the DTW algorithm is used to calculate the relative distance between the sampled data sequence and the predicted data sequence, which is used as the matching distance;

[0069] Calculate the difference between each corresponding drainage flow sampling data between the sampled data sequence and the predicted data sequence, as the data difference; calculate the mean of all data differences between the sampled data sequence and the predicted data sequence, as the first deviation coefficient;

[0070] Calculate the mean and standard deviation of all drainage flow sampling data in the sampling data sequence, and multiply them to obtain the first product; calculate the mean and standard deviation of all drainage flow sampling data in the prediction data sequence, and multiply them to obtain the second product; calculate the difference between the first product and the second product as the second deviation coefficient.

[0071] The prediction deviation of the data segment group is obtained based on the matching distance, the first deviation coefficient, and the second deviation coefficient; the matching distance, the first deviation coefficient, and the second deviation coefficient are all positively correlated with the prediction deviation.

[0072] The prediction deviation is obtained according to the formula for obtaining the prediction deviation, which is expressed as follows:

[0073]

[0074]

[0075]

[0076] Where Z represents the prediction deviation of the data segment group; Q represents the first deviation coefficient; D represents the second deviation coefficient; F k ′ This represents the k-th predicted drainage flow rate in the data segment group; F k This represents the k-th drainage flow sampling data in the data segment group; n represents the number of corresponding data between data sequences in the data segment group. The matching distance between sampled data sequence B and predicted data sequence A in the data segment group; Sd represents the mean of all predicted drainage flow data in the data segment group for the predicted data sequence; ′ This represents the standard deviation of all predicted drainage flow data in the data segment group for the predicted data sequence. Sd represents the mean of all drainage flow samples in the data segment group; Sd represents the standard deviation of all drainage flow samples in the data segment group; premnmx() represents the minimum-maximum normalization function.

[0077] In the formula for obtaining the prediction deviation, premnmx is used to... Normalize to the range [-1, 1]; This means that within the data segment group, the mean difference between all corresponding drainage flow sampling data between the sampled data sequence and the predicted data sequence is calculated to obtain the first deviation coefficient. The smaller the mean difference, the closer the first deviation coefficient is to 0, the smaller the difference between the sampled data sequence and the predicted data sequence, and the closer the prediction deviation is to 0. The smaller the matching distance, the closer it is to 0, the more similar the data between the sampled data sequence and the predicted data sequence, the smaller the difference, and the closer the prediction deviation is to 0. The difference between the product of the mean and standard deviation of all drainage flow sample data in the data segment group and the product of the mean and standard deviation of all drainage flow predicted data in the data segment group for the sampled data sequence is calculated to obtain the second deviation coefficient. This coefficient represents the deviation of the overall distribution characteristics between the sampled data sequence and the predicted data sequence in the data segment group. The closer the difference is to 0, the closer the prediction deviation is to 0, indicating that the overall level and dispersion of the data between the two data sequences in the data segment group are more similar, and the prediction deviation is closer to 0. If the prediction deviation is closer to 0, the difference between the two data sequences is smaller; conversely, if the prediction deviation is larger or smaller, the difference between the two data sequences is larger.

[0078] It should be noted that the specific DTW algorithm is a well-known technique in the field and will not be elaborated here.

[0079] It should be noted that in other embodiments of the present invention, other basic mathematical operations can also be used to construct positive and negative correlations. The specific methods are well known to those skilled in the art and will not be described in detail here.

[0080] Step S3: Based on the difference between the prediction deviation of the data segment group at the real time and the prediction deviation of other data segment groups in the historical range, obtain the relative prediction deviation at the real time; based on the relative prediction deviation at the real time and the prediction deviation of all data segment groups, adjust the drainage flow prediction data for the next time of the real time to obtain the weighted prediction data for the next time of the real time.

[0081] By understanding the prediction deviation of the data segment group at the real-time moment, we can understand the degree of deviation between the predicted data and the sampled data. By calculating the prediction deviation of historical data segment groups, we can obtain the prediction deviation level under normal circumstances and use it as a benchmark to compare the prediction level at the real-time moment. The relative prediction deviation can intuitively compare the prediction level at the real-time moment with the historical average prediction level. The larger the relative prediction deviation, the worse the prediction level at the real-time moment may be compared with the historical average prediction level, and the less accurate the prediction. Therefore, based on the difference between the prediction deviation of the data segment group at the real-time moment and the prediction deviation of other data segment groups within the historical range, we can obtain the relative prediction deviation at the real-time moment.

[0082] Preferably, in one embodiment of the present invention, the method for obtaining the relative prediction deviation includes:

[0083] Calculate the mean absolute value of the prediction deviation of all other data segment groups outside the data segment group at the real time, and use it as the average deviation value;

[0084] The difference between the absolute value of the prediction deviation and the average deviation value of the data segment group at the real-time moment is calculated and normalized to obtain the relative prediction deviation at the real-time moment. In one embodiment of the present invention, the formula for the relative prediction deviation is expressed as:

[0085]

[0086] Among them, T i Z represents the relative prediction deviation at real time i; i Z represents the prediction deviation of the data segment group at real time i; N represents the number of data segment groups; k represents the prediction deviation of the kth other data segment group; ISRU represents the nonlinear activation function.

[0087] In the formula for relative prediction deviation, It represents the mean absolute value of the prediction deviation of all other data segment groups outside the data segment group at the real time. The larger the mean value, the greater the deviation between the sampled data and the predicted data in the other data segment groups. This represents the difference between the absolute value of the prediction deviation of the data segment group at the real-time time and the mean absolute value of the prediction deviation of all other data segment groups outside the real-time data segment group. The result is mapped to the range [-1, 1] using the ISRU function. The larger the difference, the greater the deviation of the prediction deviation of the data segment group at the real-time time compared with the prediction deviation of other data segment groups. In other words, the deviation between the predicted data and the sampled data is larger, the relative prediction deviation at the real-time time is larger, and the prediction effect of the data segment group is relatively poor. Conversely, the smaller the difference, the smaller the deviation of the prediction deviation of the data segment group at the real-time time compared with the prediction deviation of other data segment groups. In other words, the deviation between the predicted data and the sampled data is smaller, the relative prediction deviation at the real-time time is smaller, and the prediction effect of the data segment group is relatively more accurate.

[0088] The relative prediction deviation at real-time reflects the magnitude of the prediction deviation of the data segment group at that real-time time compared to other data segment groups, allowing for a more accurate capture of data change trends. Prediction deviation reflects the discrepancy between the actual sampled data and the predicted data. By comprehensively considering the influence of different data segment groups, the prediction deviation of the appropriate data segment group can be more accurately selected to adjust the prediction data for the next time step, making the prediction result closer to the actual situation. Therefore, based on the relative prediction deviation at real-time and the prediction deviation of all data segment groups, the drainage flow prediction data for the next time step is adjusted to obtain the weighted prediction data for the next time step.

[0089] Preferably, in one embodiment of the present invention, the method for obtaining weighted prediction data includes:

[0090] Based on the numerical characteristics of the relative prediction deviation at real time, determine whether the drainage flow prediction data for the next time moment needs to be adjusted. If adjustment is required, obtain the weighting coefficient based on the relative prediction deviation at real time moment and the prediction deviation of all data segment groups. Calculate the product of the weighting coefficient and the corresponding drainage flow prediction data for the next time moment, and use it as the weighted prediction data for the next time moment from the real time moment.

[0091] Preferably, in one embodiment of the present invention, determining whether the predicted drainage flow rate for the next time moment needs adjustment based on the numerical characteristics of the relative prediction deviation at real-time moments includes:

[0092] If the relative prediction deviation at real time is not zero, the predicted drainage flow rate for the next time step needs to be adjusted; otherwise, no adjustment is required.

[0093] Preferably, in one embodiment of the present invention, the weighting coefficients include:

[0094] For a negative relative prediction deviation at a real time, the sum of the prediction deviation of the data segment group at the real time and the preset constant is calculated and used as a weighting coefficient.

[0095] For real-time relative prediction deviations that are positive, the smallest absolute value of the prediction deviation among all data segment groups is selected as the minimum absolute value; the sum of the minimum absolute value and a preset constant is calculated as the weighting coefficient.

[0096] Since the relative prediction deviation is obtained after normalization mapping, its range is [-1, 1]. If the relative prediction deviation at real time is negative, -1 ≤ T. i <0; if the relative prediction deviation at real time is positive, 1≥T i >0; In one embodiment of the present invention, the formula for weighted prediction data is expressed as:

[0097]

[0098] Among them, F i ′+′ 1 represents the weighted prediction data for the next time step i+1 after real-time time i; F i ′ +1 T represents the predicted drainage flow rate at the next time step i+1 after real-time time i; i Z represents the relative prediction deviation of the data segment group at real time i; i |Z| represents the prediction deviation of the data segment group at real time i; |Z| represents the absolute value of the prediction deviation of the data segment group; min() represents the minimum value function; α represents the preset constant.

[0099] In the formula for weighted prediction data, if the relative prediction deviation at real time is negative, -1 ≤ T i <0, meaning the smaller the relative prediction deviation at real time, the smaller the deviation of the prediction deviation within the data segment group compared to the prediction deviation within other data segment groups. In other words, the smaller the difference between the sampled data and the predicted data, the more reliable the prediction deviation of that data segment group can be used to adjust the drainage flow prediction data for the next time step. If the relative prediction deviation at real time is positive, 1≥T iA value greater than 0 indicates a larger relative prediction deviation at real-time. This means a greater deviation within the same data segment group compared to other data segment groups, resulting in a larger difference between the sampled and predicted data. Adjusting the drainage flow prediction data for the next time step based solely on the real-time prediction deviation would lead to excessive changes and a larger gap with the sampled drainage flow data. To avoid this excessive gap, the data segment with the smallest absolute prediction deviation value among all data segment groups should be selected for adjustment. If T... i =0, the prediction deviation of the data segment group at the real time is consistent with the prediction deviation of other data segment groups, the prediction accuracy is the same, and no adjustment is required.

[0100] It should be noted that, in one embodiment of the present invention, the preset constant is 1.

[0101] Step S4: Obtain the real-time gate opening degree at real time, and obtain the predicted gate opening degree based on the real-time drainage flow sampling data, the real-time gate opening degree, and the weighted prediction data.

[0102] The real-time gate opening degree reflects the current operating status of the drainage gate, accurately indicating the current water flow and providing crucial information for predicting the gate opening degree at the next moment. Real-time drainage flow sampling data provides real-time information on water flow; analyzing this data allows for the assessment of flow trends and prediction of gate opening requirements at the next moment. The weighted prediction value for the next moment can be analyzed based on the real-time drainage flow sampling data. By appropriately allocating weights, the gate opening degree can be adjusted more accurately, thereby optimizing drainage efficiency. Therefore, the predicted gate opening degree is obtained based on the real-time drainage flow sampling data, the real-time gate opening degree, and the weighted prediction value.

[0103] Preferably, in one embodiment of the present invention, the method for obtaining the predicted gate opening degree includes:

[0104] Obtain the maximum carrying capacity of the drainage gate; if the weighted predicted data at the next moment is less than the maximum carrying capacity, calculate the ratio of the weighted predicted data at the next moment to the drainage flow sampling data at the real time, and use it as the first ratio; calculate the product of the first ratio and the real-time gate opening degree at the real time, and use it as the predicted gate opening degree at the next moment.

[0105] If the weighted prediction data for the next moment is greater than or equal to the maximum carrying capacity flow value, the maximum gate opening degree of the drainage gate is used as the predicted gate opening degree for the next moment. In one embodiment of the present invention, the formula for predicting the gate opening degree is expressed as:

[0106]

[0107] Among them, S i+1 S represents the predicted gate opening degree at the next time step i+1 after real-time time i; i F represents the real-time gate opening degree at time i; i ′+′ 1 represents the weighted prediction data for the next time step i+1 after real-time time i; F i This represents the sampling data of drainage flow rate at real time i; F0 represents the maximum flow rate capacity of the drainage gate; ρ represents the maximum opening degree of the drainage gate.

[0108] In the formula for predicting the opening degree of the gate, if the weighted prediction data is less than the maximum flow capacity of the drainage gate, This represents the ratio of the weighted prediction data at the next time i+1 to the drainage flow sampling data at the real time i. The smaller the ratio, the smaller the weighted prediction data, and the more the gate opening degree is adjusted, the smaller the predicted gate opening degree. If the weighted prediction data is greater than or equal to the maximum flow capacity of the drainage gate, the predicted gate opening degree is fully opened to improve the overall efficiency of the drainage system.

[0109] It should be noted that, in one embodiment of the present invention, the degree of gate opening is expressed in numerical form within the range of [1, 2], with the minimum gate opening degree being 1, indicating complete closure, and the maximum gate opening degree being 2, indicating complete opening; the maximum flow capacity value can be obtained in advance by the implementer according to the specific situation of the drainage gate, which will not be elaborated here.

[0110] Step S5: Adjust the drainage gate according to the predicted gate opening degree.

[0111] The degree of gate opening directly affects the flow rate of water. By obtaining the predicted gate opening degree based on the changes in real-time predicted drainage flow sampling data, the gate opening degree can be adaptively controlled, making the drainage system operate more efficiently and reducing drainage problems caused by poor water flow or excessively high water levels, thus achieving the best drainage performance. Therefore, the drainage gate is controlled according to the predicted gate opening degree.

[0112] In the event of extreme weather, if the flow rate exceeds the maximum capacity of the drainage gate, in addition to adjusting the gate opening to its maximum extent, water can also be quickly diverted and discharged using pumping facilities, thereby achieving efficient control of drainage performance.

[0113] In summary, this invention constructs a sampled data sequence and a predicted data sequence; it obtains multiple data segment groups from the sampled data sequence and the predicted data sequence based on the volatility of the data in the sampled data sequence; it obtains the prediction deviation of each data segment group based on the difference distribution characteristics and relative distance between the two data sequences in each data segment group; it obtains the relative prediction deviation difference at the real time based on the difference between the prediction deviation of the data segment group at the real time and the prediction deviation of other data segment groups within the historical range; it obtains the weighted prediction data for the next time step; it further obtains the predicted gate opening degree; and it regulates the drainage efficiency. This invention optimizes the gate opening degree and improves drainage efficiency by obtaining accurate drainage flow prediction data for the next time step.

[0114] The present invention also proposes an intelligent control system for drainage efficiency based on drainage flow monitoring data, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any one of the steps of an intelligent control method for drainage efficiency based on drainage flow monitoring data.

[0115] An embodiment of a drainage flow data prediction method:

[0116] In existing technologies, regulating drainage efficiency in engineering projects based on drainage flow monitoring data relies on advanced data collection and analysis techniques to monitor and adjust the operating status of the drainage system in real time, thereby improving drainage efficiency. However, predicting drainage flow data for the next moment through time series analysis of drainage flow monitoring data may result in significant deviations between predicted and actual values ​​due to uncertainties such as equipment status, leading to inaccurate drainage flow prediction data. To address this issue, this embodiment provides a drainage flow data prediction method, including:

[0117] Step S1: Obtain the drainage flow sampling data sequence of the drainage gate at each sampling time in real time according to the preset sampling time period.

[0118] Step S2: Based on the variation characteristics of the drainage flow sampling data in the sampling data sequence, obtain the predicted drainage flow data at the next time step for each sampling moment, thus forming a predicted data sequence; based on the volatility of the data in the sampling data sequence, obtain multiple data segment groups between the sampling data sequence and the predicted data sequence; based on the difference distribution characteristics and relative distance between the two data sequences in each data segment group, obtain the prediction deviation of each data segment group.

[0119] Step S3: Based on the difference between the prediction deviation of the data segment group at the real time and the prediction deviation of other data segment groups in the historical range, obtain the relative prediction deviation at the real time; based on the relative prediction deviation at the real time and the prediction deviation of all data segment groups, adjust the drainage flow prediction data for the next time of the real time to obtain the weighted prediction data for the next time of the real time.

[0120] Since the specific implementation process of steps S1-S3 has been explained in detail in the above-mentioned intelligent control method for drainage efficiency based on drainage flow monitoring data, it will not be repeated here.

[0121] The technical effect of this embodiment is as follows:

[0122] This invention constructs a sampled data sequence and a predicted data sequence; based on the volatility of the data in the sampled data sequence, it obtains multiple data segment groups from the sampled data sequence and the predicted data sequence; based on the difference distribution characteristics and relative distance between the two data sequences in each data segment group, it obtains the prediction deviation of each data segment group; based on the difference between the prediction deviation of the data segment group at the real time and the prediction deviation of other data segment groups within the historical range, it obtains the relative prediction deviation difference at the real time; it obtains the weighted prediction data for the next time moment; further, it obtains the predicted gate opening degree for the next time moment; and it regulates the drainage efficiency. This invention improves the accuracy of prediction by obtaining accurate drainage flow prediction data for the next time moment.

[0123] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0124] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for intelligent control of drainage efficiency based on drainage flow monitoring data, characterized in that, The method includes: The drainage flow sampling data sequence of the drainage gate at each sampling time is obtained in real time according to the preset sampling time period; Based on the variation characteristics of the drainage flow sampling data in the sampling data sequence, the predicted drainage flow data at the next time moment for each sampling moment is obtained, forming a predicted data sequence; based on the volatility of the data in the sampling data sequence, multiple data segment groups between the sampling data sequence and the predicted data sequence are obtained; based on the difference distribution characteristics and relative distance between the two data sequences in each data segment group, the prediction deviation of each data segment group is obtained. Based on the difference between the prediction deviation of the data segment group at the real time and the prediction deviation of other data segment groups in the historical range, the relative prediction deviation at the real time is obtained; based on the relative prediction deviation at the real time and the prediction deviation of all data segment groups, the drainage flow prediction data for the next time of the real time is adjusted to obtain the weighted prediction data for the next time of the real time. Obtain the real-time gate opening degree at real time, and obtain the predicted gate opening degree based on the real-time drainage flow sampling data, the real-time gate opening degree, and the weighted prediction data; The drainage gate is adjusted according to the predicted gate opening degree; The method for obtaining the data segment group includes: Based on the volatility of the data in the sampled data sequence, the APCA method is applied to the sampled data sequence to obtain multiple data segments; the data with the smallest temporal sequence in the predicted data sequence and the sampled data sequence are aligned and arranged, and the predicted data sequence is segmented according to the segmentation position of the data segments in the sampled value sequence to obtain multiple data segments of the predicted data sequence. The data segments at corresponding positions between the sampled data sequence and the predicted data sequence are grouped into data segment groups; The method for obtaining the prediction deviation includes: In the data segment group, the DTW algorithm is used to calculate the relative distance between the sampled data sequence and the predicted data sequence, which is used as the matching distance; Calculate the difference between each corresponding drainage flow sampling data between the sampled data sequence and the predicted data sequence, as the data difference; calculate the mean of all data differences between the sampled data sequence and the predicted data sequence, as the first deviation coefficient; Calculate the mean and standard deviation of all drainage flow sampling data in the sampling data sequence, and multiply them to obtain the first product; calculate the mean and standard deviation of all drainage flow sampling data in the prediction data sequence, and multiply them to obtain the second product; calculate the difference between the first product and the second product as the second deviation coefficient. The prediction deviation of the data segment group is obtained based on the matching distance, the first deviation coefficient, and the second deviation coefficient; the matching distance, the first deviation coefficient, and the second deviation coefficient are all positively correlated with the prediction deviation. The formula for obtaining the prediction deviation is expressed as: ； ； ； in, The prediction deviation of the data segment group; Indicates the first deviation coefficient; This represents the second deviation coefficient; This indicates the predicted data sequence in the data segment group. One drainage flow forecast data; This indicates the sampling data sequence in the data segment group. One drainage flow sampling data; This indicates the number of corresponding data items between data sequences within a data segment group; For the sampled data sequence in the data segment group and predicted data sequences The matching distance between them; This represents the mean of all predicted drainage flow data in the data segment group for the predicted data sequence. This represents the standard deviation of all predicted drainage flow data in the data segment group for the predicted data sequence. This represents the mean of all drainage flow sampling data in the data segment group; This represents the standard deviation of all drainage flow sampling data in the data segment group; This represents the minimum-maximum normalization function; The method for obtaining the relative prediction deviation includes: The mean absolute value of the prediction deviation of all other data segment groups outside the data segment group at the real time is calculated as the average deviation value; Calculate the difference between the absolute value of the prediction deviation and the average deviation value of the data segment group at the real time, and perform normalization mapping to obtain the relative prediction deviation at the real time. The formula for relative prediction deviation is expressed as: ； in, Indicates real-time time The relative prediction deviation; Indicates real-time time Prediction deviation of the data segment group; Indicates the number of data segment groups; Indicates the first Prediction deviation of other data segment groups; Represents a nonlinear activation function; The method for obtaining the weighted prediction data includes: Based on the numerical characteristics of the relative prediction deviation at real time, it is determined whether the drainage flow prediction data for the next time moment needs to be adjusted. If adjustment is required, a weighting coefficient is obtained based on the relative prediction deviation at real time moment and the prediction deviation of all data segment groups. The product of the weighting coefficient and the predicted drainage flow rate at the next time step is calculated to obtain the weighted prediction data for the next time step at the current time step.

2. The intelligent control method for drainage efficiency based on drainage flow monitoring data according to claim 1, characterized in that, The method for obtaining the drainage flow prediction data includes: For the earliest sampling time in the sampling data sequence, the drainage flow sampling data of each sampling time is used as the drainage flow prediction data of the corresponding sampling time; Starting from the earliest sampling time in the time series, calculate the difference between the drainage flow sampling data at each sampling time and the corresponding previous sampling time, and use it as the first difference; The sum of the first difference and the drainage flow sampling data at each sampling time is calculated as the drainage flow prediction data for the next time step at each sampling time step.

3. The intelligent control method for drainage efficiency based on drainage flow monitoring data according to claim 1, characterized in that, The step of determining whether the drainage flow prediction data for the next time moment needs adjustment based on the numerical characteristics of the relative prediction deviation at real time includes: If the relative prediction deviation at a real-time moment is not zero, the predicted drainage flow rate for the next moment needs to be adjusted; otherwise, no adjustment is required.

4. The intelligent control method for drainage efficiency based on drainage flow monitoring data according to claim 1, characterized in that, The method for obtaining the weighting coefficients is as follows: For the relative prediction deviation at real time that is negative, the sum of the prediction deviation of the data segment group at real time and the preset constant is calculated and used as a weighting coefficient. For the relative prediction deviation at real time, which is positive, the one with the smallest absolute value of prediction deviation among all data segment groups is selected as the minimum absolute value; the sum of the minimum absolute value and the preset constant is calculated as the weighting coefficient.

5. The intelligent control method for drainage efficiency based on drainage flow monitoring data according to claim 1, characterized in that, The method for obtaining the degree of gate opening includes: Obtain the maximum carrying capacity of the drainage gate; if the weighted predicted data at the next moment is less than the maximum carrying capacity, calculate the ratio of the weighted predicted data at the next moment to the drainage flow sampling data at the real time, and use it as the first ratio; calculate the product of the first ratio and the real-time gate opening degree at the real time, and use it as the predicted gate opening degree at the next moment. If the weighted predicted data at the next moment is greater than or equal to the maximum carrying capacity flow value, the maximum gate opening degree of the drainage gate will be used as the predicted gate opening degree at the next moment.

6. A drainage efficiency intelligent control system based on drainage flow monitoring data, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent control method for drainage efficiency based on drainage flow monitoring data as described in any one of claims 1 to 5.

Images