A digital pump intelligent control method and system

Abstract

The present invention relates to the field of digital pump control technology, and in particular to a digital pump intelligent control method and system. The present invention proposes the following scheme: by acquiring real-time parameter data of a digital pump collected by a sensor, the data is input into a data analysis model including a first analysis network and a second prediction network, operation characteristics are extracted and future states are predicted, and the operation strategy of the digital pump is dynamically adjusted according to the analysis results; regional modeling is performed according to the topological structure and operation requirements of the water supply network, and related digital pump historical data is screened, and reference data is generated in combination with time window sliding processing to support operation strategies such as pressure regulation, flow optimization and overpressure protection; the present invention solves the problems of poor real-time performance, insufficient risk identification and high energy consumption in traditional water supply networks, and improves the safety, stability and operation efficiency of the water supply system.

Description

Technical Field

[0001] The present invention relates to the technical field of digital pump control, and in particular to a digital pump intelligent control method and system. Background Art

[0002] In the existing technology, the operation and management of urban water supply networks mainly rely on digital pumps to adjust pressure and flow, but traditional methods are usually based on static parameters or simple threshold judgments, which are difficult to adapt to the dynamic changes in water supply demand. During peak water use periods, traditional systems often lead to insufficient or overpressure problems due to the inability to predict water supply trends in real time; under abnormal conditions (such as leakage or equipment failure), there is a lack of intelligent analysis methods, and it is impossible to quickly identify risks and dynamically adjust strategies. In addition, the layout and configuration of digital pumps mostly rely on experience, which is difficult to accurately match with the actual operation requirements of the pipeline network, resulting in energy waste and equipment overload. Existing methods have significant deficiencies in network-wide coordinated regulation, risk identification and trend prediction, and it is difficult to meet the needs of modern water supply systems for efficient, safe and intelligent management.

[0003] In order to solve the above problems, the present application designs a digital pump intelligent control method and system. Summary of the invention

[0004] In view of the deficiencies of the prior art, the present invention provides a digital pump intelligent control method and system, which obtains the real-time parameter data of the digital pump collected by the sensor, inputs it into the data analysis model including the first analysis network and the second prediction network, extracts the operation characteristics and predicts the future state, and dynamically adjusts the digital pump operation strategy according to the analysis results. Regional modeling is performed according to the topological structure and operation requirements of the water supply network, and the historical data of the associated digital pump is screened, and reference data is generated in combination with the time window sliding process to support operation strategies such as pressure regulation, flow optimization and overpressure protection.

[0005] To achieve the above object, the present invention provides the following technical solutions: A digital pump intelligent control method is applied to a city water supply network, wherein the city water supply network includes a network node model, each node in the network node model is configured with a digital pump and a sensor component, and the method includes: Acquiring real-time parameter data of the digital pump collected by the sensor assembly; Input the real-time parameter data into a preset data analysis model, and output the data analysis results through the data analysis model, wherein the data analysis model includes a first analysis network and a second prediction network, the first analysis network is used to process the real-time parameter data, extract operation characteristics and output the current operation status, and the second prediction network outputs the prediction status according to the operation characteristics; The operation strategy of the digital pump is adjusted according to the data analysis results.

[0006] The step of acquiring the real-time parameter data of the digital pump collected by the sensor assembly comprises: According to the operation status of the water supply network, adjusting the acquisition sequence priority of the sensor components; The data collected by the sensor assembly is verified for consistency, and the data processing sequence is adjusted according to the verification result to obtain real-time parameter data, wherein the real-time parameter data includes pressure data and flow data.

[0007] The first analysis network comprises: An input layer receives the real-time parameter data according to a data processing order, performs weighted assignment on the real-time parameter data according to the receiving order, and performs channel segmentation on the weighted real-time parameter data; The feature extraction layer is used to process pressure and flow data in different channels respectively and output channel features; A feature fusion layer, which merges the channel features according to the self-attention mechanism, performs nonlinear mapping through a fully connected layer, and outputs running features, wherein the weight of the input layer is used as the initial weight matrix of the self-attention mechanism; The output layer maps the operation features to a high-dimensional space, processes them layer by layer through a pooling kernel in the high-dimensional space, and classifies the current operation status according to a Softmax classifier.

[0008] The feature extraction layer comprises: The pressure feature extraction branch extracts features from pressure data through a sparse convolutional network and calculates pressure channel features; The traffic feature extraction branch performs feature decomposition on traffic data through a group convolutional network and calculates traffic channel features.

[0009] The second prediction network comprises: An input layer, used to receive the operation characteristics and historical reference data output by the first analysis network; A trend extraction layer, used to extract pressure change trends and flow change trends, wherein the trend extraction layer includes a pressure trend path and a flow trend path; The output layer includes a pressure prediction branch and a flow prediction branch, which outputs a pressure prediction value and a flow prediction value of a future time step according to the pressure change trend and the flow change trend, wherein the future time step is calculated according to the current time.

[0010] The historical reference data is parameter data obtained by screening the historical parameter data of all digital pumps in the water supply network, and the screening includes: According to the historical parameter data of all digital pumps, the historical parameter data of digital pumps that are directly associated with the operation of the current digital pump are screened, wherein the direct operation association is calculated according to the pressure transmission path, flow interaction relationship and correlation of the historical operation data between the digital pumps; Performing time window sliding processing on the digital pump historical parameter data directly related to the operation, wherein the size of the time window is adjusted according to the current time period and the area where the current digital pump is located; The historical parameter data within the time window is used as historical reference data.

[0011] The step of adjusting the operation strategy of the digital pump according to the data analysis result includes: According to the current operating state output by the first analysis network and the predicted state output by the second prediction network, determine whether there are risks in the digital pump at the current and future time steps, wherein the risks include pressure risk, flow risk and overload risk; If there is a risk, adjust the operating strategy according to the risk type.

[0012] The adjustment operation strategy includes pressure regulation, flow optimization and overpressure protection.

[0013] The method further includes modeling and analyzing the water supply network, and configuring a digital pump at a model node of the water supply network, specifically including: Collect water supply network topology, regional characteristics data and operation requirements; Performing regional segmentation on the water supply network according to the topological structure and regional characteristic data, and outputting regional boundary conditions; The local operation characteristics are analyzed according to the boundary conditions and operation requirements of the area, and the layout position and configuration of the digital pump are output.

[0014] A digital pump intelligent control system, the system comprising a data acquisition module, a data analysis module and an operation adjustment module; The data acquisition module is used to collect real-time parameter data of the digital pump in the water supply network, including pressure data and flow data, and perform preliminary processing; The data analysis module is used to receive the real-time parameter data provided by the data acquisition module, extract the operation characteristics through the first analysis network and output the current operation status, and output the predicted values ​​of pressure and flow through the second prediction network; The operation adjustment module adjusts the operation strategy of the digital pump according to the operation status and risk prediction provided by the data analysis module, including pressure regulation, flow optimization and overpressure protection.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention combines real-time parameter data with historical reference data to accurately analyze and predict the trend of the operating status of the digital pump, and provides forward-looking operation strategy adjustments, thereby ensuring the stability and efficiency of the water supply network; the present invention combines regional characteristic data and operation requirements to optimize the layout and operation parameters of the digital pump, and also supports dynamic regulation of pressure and flow distribution, avoiding problems such as excessive energy consumption, insufficient water supply or equipment overload caused by unreasonable initial configuration or changes in operating conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 This is a flow chart of a digital pump intelligent control method according to Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the structure of a node model of a city water supply network according to Embodiment 1 of the present invention; Figure 3 This is a structural diagram of a data analysis model according to Embodiment 1 of the present invention; Figure 4 This is a first analysis network structure diagram of Example 1 of the present invention; Figure 5 This is a structural diagram of a feature extraction layer in Example 1 of the present invention; Figure 6 This is a second prediction network structure diagram of Example 1 of the present invention; Figure 7 This is a flow chart of a method for modeling and analyzing a water supply network according to Example 1 of the present invention. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0018] Example 1

[0019] See also Figure 1 The present invention provides an embodiment: a digital pump intelligent control method, the specific steps of the method are as follows: S1: Acquire real-time parameter data of the digital pump collected by the sensor component; In this embodiment, the pressure data and flow data of the digital pump are collected in real time by configuring the sensor components at the nodes of the water supply network, and the collected data are preprocessed, including dynamically adjusting the collection frequency and priority of the sensor according to the operating status of the water supply network, and performing consistency verification to ensure the accuracy and integrity of the data. For example, during peak water use periods, real-time parameter data of key nodes are collected first, and the data is weighted so that high-priority data has a greater influence weight in subsequent analysis.

[0020] S2: inputting the real-time parameter data into a preset data analysis model, and outputting data analysis results through the data analysis model; In this embodiment, the collected real-time parameter data is input into a data analysis model including a first analysis network and a second prediction network. The first analysis network processes the real-time data, extracts the local characteristics of the pressure data through a sparse convolutional network, and performs multi-dimensional feature decomposition of the flow data in combination with a grouped convolutional network to generate the operating characteristics of the current digital pump, and uses a Softmax classifier to output the current operating status, such as "normal", "overpressure warning" or "low pressure abnormality" and other status results. At the same time, the second prediction network extracts the pressure change trend and flow change trend based on the operating characteristics output by the first analysis network and the historical reference data screened from the historical parameter data of digital pumps across the entire network, and outputs the pressure and flow prediction results for multiple time steps in the future. Accurate evaluation of the current operating status of the digital pump and forward-looking prediction of future operating trends are achieved.

[0021] S3: Adjust the operation strategy of the digital pump according to the data analysis results; In this embodiment, based on the current operating status of the first analysis network and the prediction results of the second prediction network, it is determined whether the current digital pump has pressure risk, flow risk or overload risk. If there is a risk, the operation strategy of the digital pump is dynamically adjusted according to the risk type. For example, when the pressure exceeds the safety threshold, the system will automatically reduce the operating speed of the digital pump and start the pressure relief device, and at the same time notify the associated digital pumps to collaboratively share the pressure load in the high-pressure area; when insufficient flow or leakage risk is detected, the water supply flow of key nodes is prioritized, and abnormal alarm information is sent to the operation and maintenance personnel. Through multi-level response mechanisms such as dynamic pressure regulation, flow optimization and overpressure protection, the stable operation of the water supply network is ensured, while reducing the operating risk of equipment and extending its service life.

[0022] See also Figure 2, a schematic diagram of the structure of the urban water supply network node model in an embodiment of the present invention shows a node model of an urban water supply network. In this model, digital acquisition and sensor components are located in different areas (area 1, area 2 and area 3). These areas are connected through a control terminal, and the control terminal collects and transmits data from each area. Each area represents a different water supply area, and each area has sensors monitoring related data. This information is summarized by the control terminal to facilitate the effective management and control of the entire urban water supply system. In this embodiment, taking the urban water supply network as an example, a digital pump intelligent control method is proposed. By collecting the pressure and flow data of the digital pump in real time, combined with the collaborative work of the first analysis network and the second prediction network, the operating status and trend of the digital pump are accurately analyzed and predicted, so as to dynamically adjust the operating strategy of the digital pump. Although digital pumps have been widely used in water supply networks for pressure and flow control in the prior art, the prior art usually relies on a single real-time parameter data, lacks the ability to verify data quality and extract features, and is difficult to accurately identify abnormal states and predict future operating trends. For example, in a peak water supply scenario, due to a surge in user demand, the pressure in a certain area drops sharply. Existing technologies may only trigger the pump speed increase through simple threshold judgment, and fail to comprehensively consider future trends and network-wide coordination, which can easily lead to overpressure in other areas or even equipment overload. In addition, in complex scenarios with large flow fluctuations, existing technologies lack accuracy in judging abnormal conditions (such as leakage or insufficient water supply), and often require manual intervention, which increases maintenance costs and makes it difficult to respond to dynamic needs in a timely manner.

[0023] For example, in the water supply network of a new urban area, the pressure in a certain area drops below the set threshold due to concentrated water use by users. Traditional methods usually directly increase the operating speed of the digital pump in this area, which may cause overpressure problems in other areas due to unbalanced pressure distribution. This embodiment identifies the low-pressure state of the current area through the first analysis network, and predicts the pressure recovery trend through the second prediction network; at the same time, it dynamically adjusts the water supply path in combination with the historical parameter data of the adjacent digital pumps, moderately increases the pressure output in the low-pressure area, and transfers the excess load to the standby digital pump to participate in the operation. Not only did it successfully solve the low-pressure problem, but it also avoided the risk of overpressure in other areas.

[0024] Preferably, in this embodiment, the digital pump intelligent control method can also be applied to multiple practical application fields of digital pumps.

[0025] Specifically, in the field of intelligent agricultural irrigation, the method of this embodiment can intelligently control the digital pumps in the irrigation network by real-time monitoring of soil moisture and water pressure changes in different areas. For example, in arid areas, when the flow demand of a certain irrigation node surges, the system can determine the water supply status of the current node based on the operating characteristics of the first analysis network, and dynamically adjust the pump speed in combination with the trend analysis of the second prediction network to ensure the reasonable allocation of water resources. At the same time, the overpressure protection function can prevent equipment damage caused by excessive pressure in the pipe network, thereby achieving the goals of precise irrigation and water conservation.

[0026] Specifically, in the field of industrial cooling water circulation, the pressure and flow of water flow are crucial to the stable operation of equipment. The method of this embodiment can optimize the operation strategy of the cooling water pump by collecting the operating parameters (such as flow and pressure) of the cooling water pump in real time and combining it with the trend analysis of the prediction network. For example, when the cooling demand decreases, the system can automatically reduce the operating power of the cooling water pump to reduce energy consumption; when abnormal pressure in the pipe network is detected or the equipment load is too high, the system can immediately execute a dynamic adjustment strategy to share the pressure load and ensure the safe operation of the equipment.

[0027] The specific steps of S1 are as follows: S1.1: adjusting the collection order priority of the sensor components according to the operating status of the water supply network; Specifically, by giving priority to collecting data from key nodes, potential low-voltage problems can be identified in a timely manner and countermeasures can be taken. The collection order priority of sensor components is adjusted through a dynamic weight algorithm. The specific steps of the dynamic weight algorithm are as follows: S1.1.1: Calculate the real-time priority weight of each sensor node based on the operating status parameters of the water supply network (such as regional flow fluctuations, pressure gradient distribution, and historical abnormal records).

[0028] The operating status is provided by real-time monitoring data of the water supply network. For example, the peak period of the water supply network is usually accompanied by a sudden increase in regional flow and a rapid drop in pressure. Therefore, during peak periods, priority is given to the pressure and flow data at the water supply end and high-demand area nodes, while the collection frequency is temporarily reduced in low-demand areas.

[0029] S1.1.2: Optimize the collection sequence based on historical anomaly records; For example, for areas that have had frequent leaks or anomalies in the past, their sensor priority will be automatically increased to ensure that abnormal situations can be captured more quickly.

[0030] Furthermore, this dynamic adjustment is achieved through the control terminal of the water supply network. After receiving the water supply network status data, the control terminal updates the sensor's collection plan in real time according to a preset priority algorithm, so that each sensor completes the data collection task in order of priority.

[0031] S1.2: Verify the consistency of the collected data, adjust the data processing order according to the verification result, and obtain real-time parameter data, wherein the real-time parameter data includes pressure data and flow data; In this embodiment, the purpose of consistency verification is to quickly screen out sensitive data and dynamically adjust the data processing order according to the sensitivity level. Highly sensitive data will be given priority to input into the data analysis model to speed up the analysis and response to potential problems; low-sensitivity data will be processed later depending on the situation. This consistency verification strategy can not only improve the response speed of key node data, but also optimize the allocation of data processing resources, ensuring that the data analysis model focuses on important data, thereby improving the overall intelligence level of the water supply network and the efficiency of abnormal response. Especially in scenarios with peak water supply or complex flow fluctuations, this mechanism for rapid sensitive data screening significantly enhances the real-time and accuracy of the system.

[0032] Specifically, the consistency verification adopts a dynamic sensitivity analysis method, which mainly verifies and screens the data quickly from three aspects: time dimension, spatial correlation, and historical anomaly correlation.

[0033] In this embodiment, in the time dimension, the consistency of the data is verified through time series comparative analysis to determine whether the volatility of the sensor data exceeds the preset range. For example, when the pressure data of a certain sensor rises or falls continuously in a short period of time, and the amplitude of change is significantly higher than the historical average fluctuation range, the data is marked as "high sensitivity data" and given a higher priority. In this way, possible pressure anomalies in the water supply network can be quickly captured, such as sudden overpressure or low pressure risks.

[0034] In this embodiment, in the spatial correlation verification, based on the water supply network topology model, the pressure and flow data of adjacent nodes are compared to determine whether the difference between them conforms to the normal fluid dynamics distribution law. For example, if the flow of the upstream node increases sharply while the flow of the downstream node remains unchanged, it may indicate that the pipeline is blocked or leaking, and the relevant data of the node will be marked so that it will be processed first.

[0035] In this embodiment, historical anomaly correlation verification is based on data anomaly history records and trend analysis, and data of sensor nodes with frequent anomaly records in the past are preferentially marked. For example, if a node has frequently recorded low pressure during the peak water supply period, the latest data of the node will be focused on.

[0036] See also Figure 3, a structural diagram of a data analysis model according to an embodiment of the present invention, wherein the data analysis model comprises a first analysis network and a second prediction network, wherein the first analysis network is used to process the real-time parameter data, extract operation characteristics and output the current operation status, and the second prediction network outputs the prediction status according to the operation characteristics.

[0037] See also Figure 4 , the first analysis network structure diagram of the embodiment of the present invention, through the layer-by-layer collaboration of the input layer, the feature extraction layer, the feature fusion layer and the output layer, the real-time parameter data of the digital pump in the water supply network is deeply analyzed, the operation characteristics are extracted and the current operation status is evaluated. The first analysis network includes: An input layer receives the real-time parameter data according to a data processing order, performs weighted assignment on the real-time parameter data according to the receiving order, and performs channel segmentation on the weighted real-time parameter data; In this embodiment, the input layer obtains the real-time parameter data of the digital pump from the sensor component, including pressure data and flow data. Each input data stream is arranged according to a pre-arranged order. For example, among multiple nodes in the water supply network, the data of real-time input parameter 1 is sorted in the first place and will be the first input to the first analysis network, and so on, which is consistent with the sequential logic of S1.2 data processing.

[0038] In this embodiment, a unique timestamp identifier is assigned to each set of input data to record the order in which it is received. After the data is received, a time-ordered data matrix will be formed, in which each column corresponds to the real-time pressure and flow parameters of a node, and each row represents data at different time points. The initial weight of the data is directly related to the order in which it is received. The weight assignment follows the following principles: the earlier the input order, the higher the weight; the later the input order, the gradually decreasing weight. After the weighted assignment is completed, all weights are normalized to ensure that the sum of the weights is 1. The weight calculation formula is: ; in, represents the weight of the i-th received data, N represents the total number of data groups, i represents the order in which the data is received, represents normalization processing, and j represents any value from 1 to N.

[0039] Specifically, based on the weighted assignment of the input order, the input layer can automatically amplify the influence of the data that arrives first, thereby improving the efficiency and priority of data processing. In the dynamic operation scenario of the water supply network, it can ensure that real-time data with abnormal conditions can be responded to quickly.

[0040] In this embodiment, the weighted data is divided into pressure channel and flow channel according to the data type, ensuring that the pressure and flow data can enter the subsequent feature extraction path independently. The weighted pressure data is organized into a matrix form, each column corresponds to the pressure value of a different node, and each row corresponds to the pressure feature at a different time point. Similarly, the flow data is also organized into a matrix form, and each column corresponds to the flow value of a different node, which effectively reduces the interference between the pressure and flow data, makes the feature extraction more targeted, and optimizes the computational efficiency of the network.

[0041] The feature extraction layer is used to process pressure and flow data in different channels respectively and output channel features; In this embodiment, the feature extraction layer adopts a branching architecture to independently process the characteristics of pressure and flow data, including a pressure feature extraction branch and a flow feature extraction branch. Each branch uses a dedicated network structure to process its unique data characteristics.

[0042] A feature fusion layer, which merges the channel features according to the self-attention mechanism, performs nonlinear mapping through a fully connected layer, and outputs running features, wherein the weight of the input layer is used as the initial weight matrix of the self-attention mechanism; In this embodiment, the pressure channel features and flow channel features are converted into corresponding feature matrices with the same dimensions through matrix conversion. At the same time, the initial weight matrix is ​​set to the dynamic weighted assignment result from the input layer, indicating the initial importance of each feature. The feature matrices are merged through the initial weight matrix, the joint feature matrix is ​​calculated, and then the query, key and value matrices are constructed through the self-attention mechanism. Each feature in the joint feature matrix is ​​redistributed, and the weight update matrix is ​​calculated based on the distribution results. The basis for redistribution is to calculate the attention score of each feature. The weight of high-similarity features is amplified, and the weight of low-similarity features is weakened. The joint feature matrix is ​​updated through the weight update matrix.

[0043] In this embodiment, the joint feature matrix enhanced by the self-attention mechanism is input into the fully connected network for nonlinear mapping to further improve the feature expression capability. Specifically, the fully connected layer expands the low-dimensional joint features to the high-dimensional space through multi-layer mapping operations to improve the feature discrimination. Batch normalization is added after the output of each layer to ensure the stable distribution of high-dimensional features and reduce the risk of gradient disappearance or explosion. The last layer of the fully connected network compresses the high-dimensional features and outputs the running feature representation, which is directly used as the input for the output layer classification.

[0044] The output layer processes the running features layer by layer through the pooling kernel and classifies the current running status according to the Softmax classifier.

[0045] Specifically, the pooling kernel is used to process the operation features that belong to high-dimensional features layer by layer, extract important information from the features and filter out redundant features. Adaptive pooling technology is used in the pooling process to dynamically adjust the pooling strategy according to the complexity of different types of features. For example, for high-dimensional feature space, a larger pooling kernel is used to quickly reduce the dimension, while for low-dimensional feature space, a small pooling kernel is used to maintain more detailed information, thereby retaining key features and enhancing the generalization ability of the model. Finally, the operation features after pooling are input into the Softmax classifier. By calculating the probability distribution of each operation state, the category with the highest probability is preferentially selected as the final state. If the probabilities of multiple categories are similar, multiple state categories can be output at the same time and a confidence score for each category can be generated. The confidence score calculation is based on the probability output of the classifier. A comprehensive judgment report is generated based on the confidence score to accurately classify and output the current operation state.

[0046] See also Figure 5 , a structural diagram of a feature extraction layer in an embodiment of the present invention, wherein the feature extraction layer comprises: The pressure feature extraction branch extracts features from pressure data through a sparse convolutional network and calculates pressure channel features; Specifically, sparse convolutional networks effectively capture the non-uniform distribution characteristics in the data and are particularly suitable for processing local anomalies in pressure data, such as local overpressure or low pressure caused by pipeline characteristics or node layout in water supply networks.

[0047] In this embodiment, the pressure data is input in the form of a two-dimensional matrix, and each element corresponds to the pressure value of a node in the water supply network. The data is sparsely processed according to the pressure value and the set upper and lower thresholds, and only the areas with significant pressure anomalies (such as overpressure and low pressure areas) are retained, and the rest are set to zero to form a sparse matrix, and the sparse matrix is ​​convoluted by the sparse convolution kernel. The design of the convolution kernel specifically considers the spatial distribution characteristics of the pressure data. Since most of the values ​​in the sparse matrix are zero, the sparse convolution operation skips the zero value position and only calculates the non-zero elements, thereby significantly reducing the computational complexity and improving the feature extraction efficiency. At the same time, a multi-layer sparse convolution stacking structure is adopted to gradually enhance the perception of local pressure anomalies and regional pressure gradients through layer-by-layer feature extraction. After each layer, the extracted features are nonlinearly mapped using the ReLU activation function to retain key features and filter redundancy. The pressure channel feature finally output is a multidimensional feature vector, which contains: the significance of local pressure anomalies, the characteristic values ​​of regional pressure gradient distribution, and the overall pattern of pressure changes with space.

[0048] The traffic feature extraction branch performs feature decomposition on traffic data through a group convolutional network and calculates traffic channel features; Specifically, flow data has multidimensional properties, usually including spatial distribution (such as flow differences between nodes), temporal dynamics (such as fluctuation patterns of flow), and abnormal patterns (such as sudden increases or decreases). These characteristics need to be modeled independently to avoid feature aliasing. At the same time, the flow characteristics of different areas of the water supply network vary significantly. For example, there may be large fluctuations in industrial areas, while residential areas tend to be stable. After grouping the flow data, each group can focus on specific regional characteristics for analysis. The core idea of ​​grouped convolution is to divide the input tensor into multiple groups, and perform convolution operations independently in each group, so as to achieve efficient feature separation and parallel processing.

[0049] In this embodiment, the flow data is input in the form of a three-dimensional tensor, and the dimensions are , where H and W represent spatial distribution, and C represents the multi-dimensional characteristics of traffic. The input tensor is divided into multiple groups according to the channel dimension C. Each group contains several channels, representing different dimensions of traffic characteristics. The convolution operation is applied to each group separately. The convolution kernel is used to extract features within a specific dimension. The size of the convolution kernel is dynamically planned according to the size of H and W of each group. The group convolution operation can avoid interference between channel features by independently processing different dimensional characteristics, while significantly reducing computational overhead. The output of the group convolution is subjected to feature fusion through the fully connected layer, and the feature vectors of independent groups are integrated into a high-dimensional feature representation. The final output traffic channel feature is a multi-dimensional feature vector, including the spatial distribution characteristics of the traffic amplitude, the temporal correlation of traffic fluctuations, and the potential pattern of traffic anomalies.

[0050] See also Figure 6 , a second prediction network structure diagram of an embodiment of the present invention, the second prediction network includes: An input layer, used to receive the operation characteristics and historical reference data output by the first analysis network; Specifically, the operating characteristics include the pressure change rate, flow fluctuation amplitude and multidimensional feature representation extracted by the first analysis network, which are used to describe the current operating status of the digital pump; the historical reference data is obtained by screening the data directly related to the current digital pump from the historical parameter data of all digital pumps in the water supply network.

[0051] In this embodiment, the screening of historical reference data includes: According to the historical parameter data of all digital pumps, the historical parameter data of digital pumps that are directly associated with the operation of the current digital pump are screened, wherein the direct operation association is calculated according to the pressure transmission path, flow interaction relationship and correlation of the historical operation data between the digital pumps; In this embodiment, a graph model based on the topological structure of the water supply network is constructed to calculate the direct correlation between the operations of the digital pumps. The digital pump nodes in the water supply network are represented as vertices in the graph model, and the pipes are used as edges to describe the physical connection relationship between the digital pumps. At the same time, the correlation weights between the digital pumps are calculated by combining the pressure transmission path and flow interaction relationship in the historical operation data. The correlation analysis of historical parameter data is based on the following dimensions: Pressure transmission path: By analyzing the pressure distribution characteristics between nodes in the water supply network, other nodes that are highly correlated with the current digital pump pressure changes are identified. For example, when a digital pump has overpressure or underpressure, the pressure changes of the upstream or downstream nodes connected to it are directly related.

[0052] Flow interaction relationship: Based on historical flow data, the flow transfer ratio and fluctuation correlation of adjacent nodes are calculated to identify the nodes that have a greater impact on the current digital pump flow change.

[0053] Correlation of historical data: Through the correlation analysis of historical time series, select nodes that are significantly correlated with the current digital pump operation mode, such as the coordinated change trend of water supply pressure during peak hours.

[0054] Performing time window sliding processing on the digital pump historical parameter data directly related to the operation, wherein the size of the time window is adjusted according to the current time period and the area where the current digital pump is located; For example, during peak water supply periods, the time window is expanded to capture more historical data and analyze long-term operating trends; during stable operation or low demand periods, the time window is narrowed to focus only on short-term changes to improve real-time performance. The adjustment logic of the time window is also combined with the water supply characteristics of the area where the digital pump is located. For example, for the high-rise building area at the end of the water supply, the time window will focus more on capturing high-frequency fluctuation data; and for the main water supply trunk area, the window will focus more on long-term pressure stability.

[0055] The historical parameter data within the time window is used as historical reference data.

[0056] Furthermore, the input layer not only standardizes the above data to make its distribution more suitable for the processing requirements of the subsequent network, but also divides the pressure characteristics and flow characteristics into independent input paths through data channel division due to the significant differences in physical properties, timing characteristics and spatial correlation between pressure and flow data, so as to ensure that the two types of features maintain their characteristics and independence in subsequent processing.

[0057] A trend extraction layer, used to extract pressure change trends and flow change trends, wherein the trend extraction layer includes a pressure trend path and a flow trend path; In this embodiment, the pressure trend path is designed by combining a recursive graph neural network (R-GNN) and a temporal convolutional network (TCN), wherein the recursive graph neural network is used to construct the spatial correlation of pressure characteristics, dynamically modeling the pressure transmission relationship between associated digital pumps based on the topological structure of the water supply network, and the temporal convolutional network captures the long-term and short-term temporal variation of pressure characteristics through multi-layer causal convolution. Because pressure data not only has dynamic changes in time, but is also affected by the spatial distribution of associated nodes, the combination of recursive graph neural network and temporal convolutional network can simultaneously model spatiotemporal characteristics, significantly improving the recognition accuracy of pressure change trends.

[0058] In this embodiment, the flow trend path adopts a combination of a dynamic time-series graph attention network (D-TGAT) and a multi-layer residual convolutional network, in which the dynamic time-series graph attention network is used to capture the spatiotemporal dynamic characteristics of flow data, dynamically adjust the flow feature weights at different historical time points through a time-aware mechanism, and analyze the spatial correlation of flow based on the node interaction relationship of the water supply network. The residual convolutional network further models the flow characteristics in depth, focusing on the flow mutation characteristics and abnormal patterns. This structure is adopted because the flow characteristics may mutate in a short period of time (such as leakage or peak demand), and the dynamic time-series graph attention network can enhance the sensitivity to key time points, while the introduction of the residual convolutional network improves the robustness to complex flow changes.

[0059] An output layer, comprising a pressure prediction branch and a flow prediction branch, outputting a pressure prediction value and a flow prediction value of a future time step according to the pressure change trend and the flow change trend, wherein the future time step is calculated according to the current time; In this embodiment, the pressure prediction branch uses a multi-layer fully connected network (FCN) to perform nonlinear mapping on the characteristics of the pressure trend path, converts the multi-dimensional characteristics into pressure prediction results for multiple future time steps, and adds a confidence score to each time step to quantify the reliability of the prediction results. The flow prediction branch uses a recurrent neural network (RNN) to perform time series modeling on the characteristics of the flow trend path, generate a time series forecast of future flow demand, and output the flow fluctuation range to evaluate the uncertainty of flow changes. Independent branches can be optimized more accurately for their respective characteristics. The pressure prediction branch emphasizes stability and global trends, while the flow prediction branch focuses on dynamic changes and short-term fluctuations.

[0060] Exemplary, common alternative networks include long short-term memory (LSTM) and gated recurrent unit (GRU). LSTM and GRU excel in processing time series data, and are particularly suitable for capturing long- and short-term dependencies in the time dimension. However, these architectures are mainly developed for time series data. For systems with strong spatial correlations such as water supply networks, their limitation is that they cannot effectively handle the spatial relationship between digital pumps. It is easier to ignore the pressure transmission path and flow interaction characteristics between nodes, so there are deficiencies in capturing spatial dynamic correlations. This deficiency may lead to weak detection capabilities for abnormal trends caused by spatial correlations (such as overpressure or leakage in local areas) in complex water supply networks.

[0061] Furthermore, in the processing of the input data of the second prediction network, the historical parameter data has been screened, and only the data directly related to the operation of the current digital pump is retained. The screening is based on the pressure transmission path between the digital pumps, the flow interaction relationship, and the correlation of historical operation data, which reduces the interference of irrelevant data and the complexity of network calculations, while ensuring the high correlation and high quality of the input data. For example, in a zoned water supply network, only analyzing the data of neighboring nodes that directly affect the current operating status of the digital pump can effectively improve the accuracy and efficiency of modeling future trends. This screening process avoids the noise interference caused by the traditional global data input method, making the prediction results more targeted and practical.

[0062] Furthermore, the second prediction network is specially designed with a dynamic calculation mechanism for the prediction time step, so that the prediction time step can be dynamically adjusted according to the current time. For example, in a water supply network, changes in water supply demand during peak periods are usually more frequent and complex, and the prediction time step needs to be shortened appropriately to provide more real-time control support; during non-peak periods, water supply demand is relatively stable, and the time step can be appropriately extended to reduce the computational burden of the network. This dynamic adjustment strategy optimized for specific scenarios is one of the important innovations of this application in practical applications, making it perform better than traditional fixed time step prediction methods in complex water supply scenarios.

[0063] The specific steps of S3 are as follows: S3.1: judging whether there are risks in the digital pump at the current and future time steps according to the current operating state output by the first analysis network and the predicted state output by the second prediction network, wherein the risks include pressure risk, flow risk and overload risk; In this embodiment, the actual operating parameters (such as pressure, flow rate and pump speed) of the digital pump at the current time step and future trends are comprehensively analyzed to determine whether the digital pump has pressure risk, flow risk and overload risk.

[0064] Specifically, the first analysis network uses real-time parameter data to extract the local characteristics and sparse distribution of pressure data through a sparse convolutional network, and deconstructs the multi-dimensional characteristics of flow data through a grouped convolutional network, maps it into multi-dimensional operating features, and uses a Softmax classifier to classify the features into "normal", "warning", "abnormal" and other states. In this way, the first analysis network can quickly and accurately capture the current operating status of the digital pump and provide real-time operating status analysis.

[0065] Specifically, the second prediction network receives the output features and historical reference data of the first analysis network, and uses the pressure trend path and flow trend path to model the dynamic trends of pressure and flow respectively. The pressure trend path uses a recursive graph neural network to capture the pressure transfer relationship between the digital pump node and the associated nodes, and combines the time convolution network to extract the long-term and short-term change trends of pressure; the flow trend path extracts the dynamic change pattern of flow characteristics through the time series graph attention network, which is particularly suitable for capturing complex flow fluctuation characteristics. Based on the predicted future pressure and flow values, the possible risk types are comprehensively judged. For example, if the pressure in a certain period of time is predicted to exceed the threshold, the "overpressure risk" signal will be triggered, or if the flow is predicted to drop significantly, it will be marked as "insufficient flow risk".

[0066] S3.2: If risks exist, adjust the operating strategy based on the risk type, including pressure regulation, flow optimization, and overpressure protection.

[0067] Specifically, when an overpressure risk is detected, the system reduces the pressure output by lowering the speed of the digital pump and starts the collaborative sharing of pressure by the associated digital pumps. To achieve this function, the system dynamically optimizes the pressure distribution in the water supply area based on the predicted pressure change trend. For example, the topological relationship of the water supply network is modeled through graph convolution, and the pressure transmission path between the nodes of the network is adjusted to ensure that the load in the high-pressure area is transferred to the low-pressure area to avoid overloading of single-node equipment. In addition, when necessary, the system starts the pressure relief device to release excess pressure in the network, and transmits the alarm signal of overpressure risk to the operation and maintenance end through the Internet of Things platform, further enhancing the safety of the water supply network.

[0068] For flow risks, when the system determines that the flow is insufficient or the leakage trend is significant, it first dynamically optimizes the flow output capacity of the digital pump, such as adding a multi-scale residual module to the flow prediction path, enhancing the response capability to short-term flow anomalies, and increasing the pump speed to increase the flow; secondly, it reallocates the water supply path based on the time series prediction results, giving priority to ensuring the water supply flow in high-demand areas, such as by adjusting the speed of the digital pump in the adjacent area to supplement the water volume in the low-flow area; in addition, in the case of leakage risk, it locates possible leakage areas based on the pressure change data, reduces losses by reducing the water supply pressure in the leakage area, and sends an alarm signal to the operation and maintenance end.

[0069] In response to the overload risk, when the system determines that the digital pump has been under long-term high load or the operating parameters are close to the equipment safety threshold, it will give priority to reducing the operating load of the digital pump and let the backup digital pump take over some tasks to reduce the pressure of the main pump. In addition, the system adjusts the target operating parameters of the digital pump based on the future trend results of the second prediction network to prevent equipment wear and efficiency reduction caused by future high-load operation in advance.

[0070] See also Figure 7 , a flow chart of a modeling and analysis method for a water supply network according to an embodiment of the present invention, this embodiment also provides a modeling and analysis method for a water supply network, and the specific steps of the method are as follows: S401: Collect the topological structure, regional characteristic data and operation requirements of the water supply network; In this embodiment, the topological structure of the water supply network includes the node distribution, length, diameter, material and connection relationship of the pipeline, which are collected in real time through the geographic information system (GIS) and sensor network deployed in the network. Regional characteristic data include terrain height difference, user distribution density, water demand type (such as industrial area, residential area, commercial area), which are obtained through terrain modeling and water demand survey. At the same time, the operation demand is determined by the real-time monitored flow, pressure change and seasonal fluctuation data. This multi-source data collection and fusion provides a comprehensive basic input for subsequent analysis. By introducing a dynamic data collection module, the data of key nodes can be collected first according to the importance of the nodes. For example, in areas with large height differences, the system prioritizes the collection of pressure and flow data of high points and terminal nodes to capture possible water supply bottlenecks. Through this collection method, it can be ensured that the dynamic changes of the pipeline network can be accurately reflected, laying a data foundation for water supply optimization.

[0071] S402: Segment the water supply network into regions according to the topological structure and regional characteristic data, and output regional boundary conditions; The core of regional segmentation is to divide the water supply network into several sub-regions with independent characteristics, each of which is relatively independent in operation but cooperates as a whole. In order to achieve this goal, this embodiment adopts a regional segmentation method based on a graph segmentation algorithm. By constructing a graph model of the water supply network, the nodes of the network are regarded as vertices and the pipes are regarded as edges, and on this basis, the partition is performed by the minimum cut algorithm or the K-means clustering method. The regional boundary conditions are determined by calculating the pressure gradient and flow transfer relationship of the internal and external nodes of the region, and the inlet pressure, outlet pressure and key nodes of the region are output. Through regional segmentation, different characteristic regions such as height difference areas, flow fluctuation areas and terminal areas can be effectively isolated, thereby providing a more targeted basis for the layout of digital pumps. For example, in a complex terrain water supply network including mountains and plains, the height difference areas separated can be independently set with digital pumps to adjust the pressure, while reducing the interference of high pressure on low pressure areas. In this way, the water supply demand in the region can be met independently, and global coordination is guaranteed by boundary conditions.

[0072] S403: Analyze local operation characteristics according to the regional boundary conditions and operation requirements, and output the layout position and configuration of the digital pump; After completing the regional segmentation and determining the boundary conditions, the local operating characteristics are further analyzed according to the operating requirements of the region, and the layout position and configuration parameters of the digital pump are finally determined. This embodiment adopts the pressure-flow coupling analysis technology to calculate the pump head and flow range required for each key node according to the average pressure demand and peak flow demand in the region. Key nodes include the entrance node of the high-demand area, the terminal water supply node, and the node with significant pressure attenuation in the pipe network. For example, by comparing the flow demand fluctuations and pressure stability of each node in the region, the node with the lowest pressure is preferentially selected as the installation location of the digital pump. In addition, this embodiment adopts a pump configuration optimization algorithm based on historical operating data, and dynamically adjusts the operating parameters of the digital pump, such as head, speed and power, according to the water consumption fluctuation curve of the node, to ensure that the digital pump can operate in the high-efficiency range. In this way, the installation of the digital pump can not only meet the operating needs in the region, but also adapt to the coordination goals of the entire network, thereby achieving global energy saving and regional water supply stability.

[0073] In this embodiment, the modeling and analysis method of the water supply network serves as a prerequisite for this application, and is intended to provide an accurate optimization basis for the subsequent optimization of the digital pump operation strategy.

[0074] Specifically, by collecting the topological structure, regional characteristic data and operation requirements of the water supply network, accurately dividing the network area and determining the boundary conditions, and further analyzing the layout and configuration parameters of the digital pump in combination with the regional operation characteristics, it is ensured that the initial layout and operation strategy of the digital pump can be highly matched with the dynamic requirements of the actual network. It can achieve refined zoning and efficient resource allocation of the water supply network, thereby avoiding pressure distribution imbalance, flow waste and equipment overload problems caused by unreasonable layout or inaccurate initial conditions, and improving the overall efficiency and stability of the water supply system.

[0075] Furthermore, the network node model in this embodiment can be applied to the dynamic monitoring and optimization of key nodes in the water supply network. In peak water consumption scenarios, the node model can identify water supply bottlenecks and potential pressure imbalance areas in real time as a basis for dynamically adjusting the digital pump operation strategy; in low-demand periods, the node model can support digital pump power optimization and reduce energy consumption.

[0076] Example 2

[0077] The present invention provides an embodiment: a digital pump intelligent control system, the system comprising a data acquisition module, a data analysis module and an operation adjustment module; The data acquisition module is used to collect real-time parameter data of the digital pump in the water supply network, including pressure data and flow data, and perform preliminary processing; The data analysis module is used to receive the real-time parameter data provided by the data acquisition module, extract the operation characteristics through the first analysis network and output the current operation status, and output the predicted values ​​of pressure and flow through the second prediction network; The operation adjustment module adjusts the operation strategy of the digital pump according to the operation status and risk prediction provided by the data analysis module, including pressure regulation, flow optimization and overpressure protection.

[0078] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A digital pump intelligent control method is applied to a city water supply network, wherein the city water supply network is configured with a network node model, and each node in the network node model is configured with a digital pump and a sensor component, characterized in that: The method comprises: Acquiring real-time parameter data of the digital pump collected by the sensor assembly; Input the real-time parameter data into a preset data analysis model, and output the data analysis results through the data analysis model, wherein the data analysis model includes a first analysis network and a second prediction network, the first analysis network is used to process the real-time parameter data, extract operation characteristics and output the current operation status, and the second prediction network outputs the prediction status according to the operation characteristics; The operation strategy of the digital pump is adjusted according to the data analysis results.

2. According to claim 1, a digital pump intelligent control method is characterized in that: The step of acquiring the real-time parameter data of the digital pump collected by the sensor assembly comprises: According to the operation status of the water supply network, adjusting the acquisition sequence priority of the sensor components; The data collected by the sensor assembly is verified for consistency, and the data processing sequence is adjusted according to the verification result to obtain real-time parameter data, wherein the real-time parameter data includes pressure data and flow data.

3. According to claim 2, a digital pump intelligent control method is characterized in that: The first analysis network comprises: An input layer receives the real-time parameter data according to a data processing order, performs weighted assignment on the real-time parameter data according to the receiving order, and performs channel segmentation on the weighted real-time parameter data; The feature extraction layer is used to process pressure and flow data in different channels respectively and output channel features; A feature fusion layer, which merges the channel features according to the self-attention mechanism, performs nonlinear mapping through a fully connected layer, and outputs running features, wherein the weight of the input layer is used as the initial weight matrix of the self-attention mechanism; The output layer maps the operation features to a high-dimensional space, processes them layer by layer through a pooling kernel in the high-dimensional space, and classifies the current operation status according to a Softmax classifier.

4. According to claim 3, a digital pump intelligent control method is characterized in that: The feature extraction layer comprises: The pressure feature extraction branch extracts features from pressure data through a sparse convolutional network and calculates pressure channel features; The traffic feature extraction branch performs feature decomposition on traffic data through a group convolutional network and calculates traffic channel features.

5. According to claim 1, a digital pump intelligent control method is characterized in that: The second prediction network comprises: An input layer, used to receive the operation characteristics and historical reference data output by the first analysis network; A trend extraction layer, used to extract pressure change trends and flow change trends, wherein the trend extraction layer includes a pressure trend path and a flow trend path; The output layer includes a pressure prediction branch and a flow prediction branch, which outputs a pressure prediction value and a flow prediction value of a future time step according to the pressure change trend and the flow change trend, wherein the future time step is calculated according to the current time.

6. A digital pump intelligent control method according to claim 5, characterized in that: The historical reference data is parameter data obtained by screening the historical parameter data of all digital pumps in the water supply network, and the screening includes: According to the historical parameter data of all digital pumps, the historical parameter data of digital pumps that are directly associated with the operation of the current digital pump are screened, wherein the direct operation association is calculated according to the pressure transmission path, flow interaction relationship and correlation of the historical operation data between the digital pumps; Performing time window sliding processing on the digital pump historical parameter data directly related to the operation, wherein the size of the time window is adjusted according to the current time period and the area where the current digital pump is located; The historical parameter data within the time window is used as historical reference data.

7. A digital pump intelligent control method according to claim 1, characterized in that: The step of adjusting the operation strategy of the digital pump according to the data analysis result includes: According to the current operating state output by the first analysis network and the predicted state output by the second prediction network, determine whether there are risks in the digital pump at the current and future time steps, wherein the risks include pressure risk, flow risk and overload risk; If there is a risk, adjust the operating strategy according to the risk type.

8. A digital pump intelligent control method according to claim 7, characterized in that: The adjustment operation strategy includes pressure regulation, flow optimization and overpressure protection.

9. A digital pump intelligent control method according to claim 1, characterized in that: The method further includes modeling and analyzing the water supply network, configuring a digital pump at a node of the network node model, and modeling and analyzing the water supply network includes: Collect water supply network topology, regional characteristics data and operation requirements; Performing regional segmentation on the water supply network according to the topological structure and regional characteristic data, and outputting regional boundary conditions; The local operation characteristics are analyzed according to the boundary conditions and operation requirements of the area, and the layout position and configuration of the digital pump are output.

10. A digital pump intelligent control system, used to implement a digital pump intelligent control method according to any one of claims 1 to 9, characterized in that: The system includes a data acquisition module, a data analysis module and an operation adjustment module; The data acquisition module is used to collect real-time parameter data of the digital pump in the water supply network, including pressure data and flow data, and perform preliminary processing; The data analysis module is used to receive the real-time parameter data provided by the data acquisition module, extract the operation characteristics through the first analysis network and output the current operation status, and output the predicted values ​​of pressure and flow through the second prediction network; The operation adjustment module adjusts the operation strategy of the digital pump according to the operation status and risk prediction provided by the data analysis module, including pressure regulation, flow optimization and overpressure protection.

Images