CN-122027457-A - Multi-node synchronous monitoring system and method for early fault warning of sprinkler

CN122027457ACN 122027457 ACN122027457 ACN 122027457ACN-122027457-A

Abstract

The invention provides a multi-node synchronous monitoring system for early fault warning of a sprinkler, which belongs to the technical field of computer data processing and is used for accurately capturing early weak faults, remarkably reducing false alarm rate and realizing rapid positioning of faults by means of combination of multi-node synchronous data acquisition and association mapping, preliminary abnormal-based dynamic weight focusing and reaction verification. According to the invention, through the synergistic effect of dynamic weight adjustment and reaction verification, the identification precision and alarm reliability of early weak faults are improved, the acquisition precision of corresponding associated data obtained by constructing a monitoring node group is constructed, the calculation and communication resources are dynamically inclined to the potential fault area only when the initial abnormality is identified, the feedback mechanism can timely release the resources confirmed to be free of abnormal areas, a wider monitoring range and a longer operation period can be supported under the same hardware configuration, the faults are locked in the physical area between the associated monitoring nodes, and the manual investigation time is shortened.

Inventors

TAO JIANPING
ZHANG SHIRUI
LIU JIAYU
CHEN BAOFA
XIA HAO
GENG WENJING

Assignees

江苏省农业科学院

Dates

Publication Date: 20260512
Application Date: 20251114

Claims (10)

1. A multi-node synchronous monitoring method for early-stage fault warning of a sprinkler, the method being applied to a multi-node synchronous monitoring system for early-stage fault warning of a sprinkler, the method being applied to a processing terminal, the method comprising: Acquiring sensing terminal data of a plurality of first sensing terminals arranged on a sprinkler and a plurality of second sensing terminals arranged on a region to be sprinkler-sprayed, and establishing an association information mapping relation based on the sensing terminal data and a historical timestamp, wherein the association information mapping relation is based on a sensing terminal specific node output and reflects a sensing terminal node set with association topological relation with the specific node; When the processing terminal identifies that M specific nodes are abnormal, N associated monitoring nodes are obtained based on the association information mapping relation, the number of N is larger than M, the N associated monitoring nodes and the M specific nodes form a monitoring node group, and the number of second sensing terminals in the N associated monitoring nodes is larger than the number of first sensing terminals so as to increase the in-situ monitoring frequency of the monitoring node group; Judging whether the association topological relation of the monitoring node group exists or not, if so, generating a fault alarm signal or executing false alarm elimination operation, and if no association is established, eliminating abnormal operation.
2. The multi-node synchronization monitoring method for early failure warning of a sprinkler according to claim 1, wherein the establishing of the association information mapping relation based on the sensing terminal data and the history timestamp specifically comprises: Acquiring sensing terminal data of a sensing terminal, wherein the sensing terminal data comprises historical timestamp data, geographic position information data and sensing reading of the sensing terminal; Constructing a graph structure based on the data of each induction terminal, wherein each induction terminal corresponds to one vertex, and the edges between any two vertices are established according to the data similarity; The data similarity is determined by calculating the reading difference value and the space distance of the two sensing terminals in a preset time window; screening the central node as a specific node based on the relevance weight threshold; and starting from the specific node, traversing the nodes connected with the specific node, and generating the associated information mapping relation.
3. The multi-node synchronous monitoring method for early failure warning of sprinkler according to claim 2, characterized in that, The forming the N associated monitoring nodes and M specific nodes into the monitoring node group comprises the following steps: Comparing the sensing terminal data reading obtained in real time with the association information mapping relation to identify N association monitoring nodes deviating from the model; Querying the N association monitoring node association information mapping tables to determine associated M specific nodes which have logical association with the N association monitoring nodes; And improving the data acquisition frequency and analysis depth of the N associated monitoring nodes and the M associated specific nodes to form the monitoring node group.
4. The multi-node synchronization monitoring method for early-stage fault alerting of a sprinkler according to claim 3, wherein the enhancing of the data collection frequency and the parsing depth of the N associated monitoring nodes and the associated M specific nodes to form the monitoring node group comprises: Acquiring a historical fault database and matching candidate association features corresponding to the association monitoring node types from the historical fault database, wherein the historical fault database stores predefined association features, the association features comprise a first association feature serving as a trigger condition and a second association feature serving as an expected response, and a time constraint, namely a preset time window is attached to the association features; performing topology structure screening on the candidate associated features by using the associated information mapping relation to generate an associated feature chain; and verifying whether a second associated feature in the associated feature chain is triggered within a time window after the first associated feature occurs based on the high-frequency data of the monitoring node group.
5. The multi-node synchronization monitoring method for sprinkler early failure warning according to claim 4, further comprising, after the step of verifying whether a second associated feature in the chain of associated features is triggered: Continuously monitoring the sensing terminal data of the first association feature when the second association feature is not triggered; triggering the false alarm elimination operation when the sensing terminal data returns to the range of the basic operation model; And dynamically adjusting the association strength parameters between corresponding nodes in the association information mapping relation according to the result of the false alarm elimination operation.
6. The multi-node synchronization monitoring method for sprinkler early failure warning according to claim 4, further comprising, after the step of verifying whether a second associated feature in the chain of associated features is triggered: when the second association feature is triggered, generating a quantitative characterization parameter according to the response relation between the first association feature and the second association feature; updating the verification priority of the associated feature chain based on the quantitative characterization parameters; and adjusting the information acquisition frequency strategy of the specific node and the associated monitoring node by utilizing the quantitative characterization parameters.
7. The multi-node synchronization monitoring method for early warning of a sprinkler according to claim 6, wherein the step of utilizing the quantitative characterization parameter to strengthen the information collection frequency policy of the subsystem in which the specific node and the associated monitoring node are located comprises: Determining a target subsystem needing intensive monitoring according to the quantitative characterization parameters; Adjusting a judgment threshold value of the basic operation model of the node in the target subsystem so as to improve monitoring sensitivity; And synchronizing the adjusted judging threshold value of the basic operation model to the association information mapping relation.
8. The multi-node synchronous monitoring method for early-warning of faults in a sprinkler of claim 1 in which the step of generating a fault warning signal further comprises: positioning the physical positions of the specific node and the association monitoring node based on the association information mapping relation to determine a fault section; integrating the verification information of the fault section and the associated feature chain to generate a maintenance guidance report; and sending an automatic isolation instruction to a control unit at the upstream of the fault section according to the maintenance guidance report.
9. The multi-node synchronization monitoring method for early failure warning of a sprinkler according to claim 1, wherein the association topology further comprises the steps of: Analyzing the sensing terminal data of all nodes in the monitoring node group to extract multidimensional cooperative change characteristics; performing fusion judgment on the multidimensional cooperative change feature and the associated feature chain to generate a comprehensive confidence score; and adjusting the alarm level of the fault alarm signal according to the comprehensive confidence score.
10. A multi-node synchronous monitoring system for early-stage fault warning of a sprinkler, applied to the multi-node synchronous monitoring method for early-stage fault warning of a sprinkler according to any one of claims 1 to 9, characterized in that the system is configured to include: The data acquisition and mapping module is used for acquiring the time stamp synchronous data and the geographic position information of a plurality of nodes and establishing an associated information mapping relation; The dynamic focusing module is used for identifying specific nodes based on the association information mapping relation and increasing data acquisition weight to form a monitoring node group; the association topological relation verification module is used for applying association topological relation to the monitoring node group, generating a fault alarm signal or executing false alarm elimination operation by verifying an association characteristic chain; and the self-optimizing module is used for dynamically adjusting parameters and information acquisition strategies in the association information mapping relation according to the execution result of the association topological relation.

Description

Multi-node synchronous monitoring system and method for early fault warning of sprinkler Technical Field The invention relates to the technical field of computer data processing, in particular to a multi-node synchronous monitoring system for early fault warning of a sprinkler. Background The large-scale sprinkling machine is widely applied to agricultural irrigation, and plays a vital role in water conservancy management and irrigation operation of a large-area farmland. Through the efficient water resource utilization, the large sprinkler can ensure that crops are supplied with sufficient water in drought seasons, so that the agricultural productivity and the crop yield are improved. In addition, the sprinkling irrigation technology has obvious advantages in the aspects of water resource distribution, water saving and labor force reduction compared with the traditional irrigation mode. However, once the sprinkler fails, the irrigation system may be interrupted, which further affects the normal growth of crops, resulting in large-area water resource waste and crop yield reduction, and even economic loss. In order to avoid negative effects caused by faults of a sprinkler, various fault detection methods have been proposed in the industry, and more sensors need to be deployed and the cost is high due to diversity, and because the sprinkler can generate instantaneous data fluctuation such as water hammer effect when starting, stopping or switching in normal working conditions, the normal fluctuation is easily misjudged as faults by a simple threshold judgment method, so that the misinformation rate of a system is high, data errors exist, the response instantaneity is affected, the fault discovery is delayed, and the optimal maintenance time is missed, so that a more efficient and accurate detection technology is needed. Disclosure of Invention The embodiment of the invention provides a multi-node synchronous monitoring system for early fault warning of a sprinkler, which is used for accurately capturing early weak faults, remarkably reducing false alarm rate and realizing rapid positioning of faults by means of combination of multi-node synchronous data acquisition and association mapping, dynamic weight focusing based on preliminary abnormality and reaction verification. In order to achieve the above purpose, the invention adopts the following technical scheme: In a first aspect, a multi-node synchronous monitoring method for early failure warning of a sprinkler is provided, the method comprising: Acquiring sensing terminal data of a plurality of first sensing terminals arranged on a sprinkler and a plurality of second sensing terminals arranged on a region to be sprinkler-sprayed, and establishing an association information mapping relation based on the sensing terminal data and a historical timestamp, wherein the association information mapping relation is based on a sensing terminal specific node output and reflects a sensing terminal node set with association topological relation with the specific node; When the processing terminal identifies that M specific nodes are abnormal, N associated monitoring nodes are obtained based on the association information mapping relation, the number of N is larger than M, the N associated monitoring nodes and the M specific nodes form a monitoring node group, and the number of second sensing terminals in the N associated monitoring nodes is larger than the number of first sensing terminals so as to increase the in-situ monitoring frequency of the monitoring node group; Judging whether the association topological relation of the monitoring node group exists or not, if so, generating a fault alarm signal or executing false alarm elimination operation, and if no association is established, eliminating abnormal operation. Optionally, establishing the association information mapping relationship based on the sensing terminal data and the historical timestamp specifically includes: Acquiring sensing terminal data of a sensing terminal, wherein the sensing terminal data comprises historical timestamp data, geographic position information data and sensing reading of the sensing terminal; Constructing a graph structure based on the data of each induction terminal, wherein each induction terminal corresponds to one vertex, and the edges between any two vertices are established according to the data similarity; The data similarity is determined by calculating the reading difference value and the space distance of the two sensing terminals in a preset time window; screening the central node as a specific node based on the relevance weight threshold; and starting from the specific node, traversing the nodes connected with the specific node, and generating the associated information mapping relation. Optionally, the forming the N associated monitoring nodes and the M specific nodes into the monitoring node group includes: Comparing the sensing terminal data reading obtained in real