Intelligent alarm diagnosis and intelligent supervision method for pump station

Abstract

The invention discloses an intelligent alarm diagnosis and intelligent inventory monitoring method for a pump station, and the method specifically comprises the following steps: 1, carrying out the preliminary alarm diagnosis, and carrying out the diagnosis through a built-in algorithm model; step 2, early warning analysis: further analysis is carried out by using an early warning model on the basis of the preliminary diagnosis; step 3, model construction: selecting components from the component library through dragging, and constructing a visual model according to alarm logic connection; 4, model editing and testing: during editing, the progress can be stored at any time, a testing function can be provided, a test operation model verification logic and a triggering condition can be clicked, and a result is displayed below a page, so that adjustment and optimization are facilitated; and step 5, model management: a management page displays all models, supports editing operation, improves the alarm diagnosis efficiency and accuracy, constructs a visual alarm diagnosis model through a drag-and-drop component, and enables a user to quickly define and configure alarm logic without compiling complex codes, so that the technical threshold is reduced, and the model construction efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of pump station alarm technology, specifically to a method for intelligent alarm diagnosis and smart monitoring of pump stations. Background Technology

[0002] Integrated alarm for a pump station group refers to a comprehensive alarm mechanism integrated into a system consisting of multiple pump stations. This mechanism monitors and detects the operating status, equipment parameters, and environmental conditions of each pump station, promptly identifying abnormalities and triggering alarms. This allows management personnel to take swift action to ensure the safe and stable operation of the pump station group.

[0003] In the operation and maintenance management of pumping stations, the monitoring and processing of alarm information is a crucial link. Traditional alarm systems often only provide simple threshold alarm functions, which cannot meet the needs of comprehensive monitoring of the status of pumping station equipment. In addition, traditional alarm systems lack a visual modeling process, making it difficult for users to build complex alarm models, which seriously affects the accuracy and timeliness of alarm information. Summary of the Invention

[0004] The purpose of this invention is to provide a method for intelligent alarm diagnosis and smart monitoring of pump stations that uses drag-and-drop components to build a visual alarm diagnosis model. This method allows users to quickly define and configure alarm logic without writing complex code, thus lowering the technical threshold and improving the efficiency of model building. It can solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for intelligent alarm diagnosis and smart monitoring of pump stations, specifically including the following steps: Step 1: Preliminary alarm diagnosis. In the preliminary alarm diagnosis stage, the built-in algorithm model is mainly used for alarm diagnosis. Step 2: Early warning analysis. Based on the initial alarm diagnosis, further early warning analysis is conducted using the early warning model. Step 3: Model Building. In the model building stage, select the required components from the component library by dragging and dropping, and connect the components according to the alarm logic flow to build a visual alarm diagnosis model. The connections between components represent data flow or logical dependencies, making the model structure clear and easy to understand. At the same time, configure component parameters such as alarm threshold, alarm level and classification and alarm information to meet actual needs. Step 4: Model Editing and Testing. During the model editing process, you can click the save button at any time to save the current progress and ensure that the work is not lost due to unexpected interruptions. At the same time, a model testing function is provided. Click the test button to run the current model to verify the correctness of the logic and whether the alarm triggering conditions meet the expectations. The test results will be displayed at the bottom of the page, including the triggering status, intermediate calculation results and final output, which facilitates the adjustment and optimization of the model. Step 5: Model Management. The model management page displays all created models and supports editing, copying, enabling, disabling, and deleting models. Enabled models will be automatically applied to the real-time data stream for alarm diagnosis and analysis, enabling real-time monitoring and early warning of the pump station group's operating status. Through the model management function, the model library can be easily managed and optimized, improving the efficiency and accuracy of alarm diagnosis.

[0006] Preferably, the types of models in step one include threshold alarm, rate alarm, dead point alarm, jitter alarm, and multi-point deviation alarm. Threshold alarm: When the monitored measurement value exceeds the preset upper and lower limits, a threshold alarm will be triggered. This alarm method is simple and direct and can quickly identify abnormal values ​​that exceed the normal range. Rate alarm: By monitoring the rate of data change, a rate alarm will be triggered when the rate of change exceeds a set threshold. This method can capture the trend of data change and detect potential anomalies in a timely manner. Dead spot alarm: The dead spot alarm focuses on the situation where the data hardly changes within a set time window. When the data remains constant or changes very little for a long time, it may mean that the equipment is malfunctioning or the sensor is not working. At this time, the dead spot alarm will be triggered. Jitter alarm: The jitter alarm focuses on situations where data fluctuates frequently within a short period of time. When the data fluctuation exceeds the normal range, it may mean that the device is unstable or subject to external interference, which will trigger the jitter alarm. Multi-point deviation alarm: By comparing the deviation of a single measuring point from the overall level of multiple measuring points, a multi-point deviation alarm will be triggered when the deviation exceeds a set threshold. This method can identify local anomalies and improve the accuracy of the alarm.

[0007] Preferably, the early warning model in step two includes time-limited early warning, residual early warning, and probability early warning; Time Limit Warning: Monitors whether a task or process is completed within the specified time. If a task or process is not completed within the time limit, a time limit warning will be triggered to remind relevant personnel to take timely measures. Residual warning: Based on the prediction model, when the residual between the actual value and the model prediction value exceeds the set threshold, a residual warning will be triggered. This method can predict future abnormal trends and take measures in advance to avoid failures. Probability warning: By calculating the probability of an event occurring through a probability model, a probability warning will be triggered when the probability exceeds a set threshold. This method can quantify risk and provide a scientific basis for decision-making.

[0008] Compared with the prior art, the beneficial effects of the present invention are: I. Improve the efficiency and accuracy of alarm diagnosis. By building a visual alarm diagnosis model through drag-and-drop components, users can quickly define and configure alarm logic without writing complex code. This not only lowers the technical threshold but also improves the efficiency of building alarm diagnosis models. At the same time, the model testing function ensures the correctness of the logic and the accuracy of the alarm triggering conditions, further improving the reliability of alarm diagnosis. Second, it supports multiple alarm algorithm models. The system has preset quality alarms, threshold alarms, rate alarms, dead point alarms, jitter alarms and deviation alarms, covering a variety of possible alarm scenarios. These models can be selected and configured according to actual needs to meet the monitoring needs of different devices and ensure the timeliness and accuracy of alarms.

[0009] Third, the system enables intelligent early warning and diagnosis. It supports time-limited early warning, residual early warning, and probability early warning. Through model training and analysis, it can predict possible alarm situations and trigger early warnings in advance. At the same time, the diagnostic modeling function can set alarm triggering conditions according to the flowchart pattern and provide explanations of alarm causes and handling suggestions to help users quickly locate problems and take corresponding measures.

[0010] IV. Optimized user experience and ease of operation: The system interface is simple and clear, with a reasonable functional layout, allowing users to easily get started. Real-time alarm and historical alarm pages provide detailed alarm information and operation options, facilitating user viewing and handling of alarms. The access control function ensures system security and data confidentiality, while role-based permission allocation improves system flexibility and controllability.

[0011] Fifth, the system enhances the operation and maintenance management of pumping stations. Through real-time monitoring and alarm diagnosis, it promptly detects and handles equipment failures, reducing downtime and maintenance costs. Simultaneously, the alarm information and diagnostic suggestions provided by the system offer strong decision support to maintenance personnel, improving operational efficiency and equipment reliability. Management personnel can comprehensively understand the pumping station's operation and maintenance status through the system, optimize resource allocation, and ensure the safe and stable operation of the pumping station. Attached Figure Description

[0012] Figure 1 This is a system architecture diagram of the present invention; Figure 2 This is a schematic diagram of the measurement point management interface of the present invention; Figure 3 This is a schematic diagram of the overall measurement point interface of the present invention; Figure 4 This is a schematic diagram of the measurement point trend query page of the present invention; Figure 5 This is a schematic diagram of the real-time alarm monitoring page of the present invention; Figure 6 This is a schematic diagram of the fault diagnosis analysis report of the present invention. Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] Please see Figures 1 to 6 The diagram illustrates a method for intelligent alarm diagnosis and smart monitoring of pump stations, which specifically includes the following steps: Step 1: Preliminary alarm diagnosis. In the preliminary alarm diagnosis stage, the built-in algorithm model is mainly used for alarm diagnosis. Step 2: Early warning analysis. Based on the initial alarm diagnosis, further early warning analysis is conducted using the early warning model. Step 3: Model Building. In the model building stage, select the required components from the component library by dragging and dropping, and connect the components according to the alarm logic flow to build a visual alarm diagnosis model. The connections between components represent data flow or logical dependencies, making the model structure clear and easy to understand. At the same time, configure component parameters such as alarm threshold, alarm level, classification and alarm information to meet actual needs. Step 4: Model Editing and Testing. During the model editing process, you can click the save button at any time to save the current progress and ensure that the work is not lost due to unexpected interruptions. At the same time, a model testing function is provided. Click the test button to run the current model to verify the correctness of the logic and whether the alarm triggering conditions meet the expectations. The test results will be displayed at the bottom of the page, including the triggering status, intermediate calculation results and final output, which facilitates the adjustment and optimization of the model. Step 5: Model Management. The model management page displays all created models and supports editing, copying, enabling, disabling, and deleting them. Enabled models will be automatically applied to real-time data streams for alarm diagnosis and analysis, enabling real-time monitoring and early warning of the pump station group's operating status. The model management function allows for convenient management and optimization of the model library, improving the efficiency and accuracy of alarm diagnosis. Efficient and accurate alarm diagnosis is achieved through five core steps: First, in the preliminary alarm diagnosis stage, the built-in algorithm model is used to initially screen and judge alarm information. Next, in the early warning analysis stage, based on the preliminary diagnosis results, the early warning model further analyzes potential alarm risks. In the model building stage, users can select required components from the component library by dragging and dropping, building a visual alarm diagnosis model according to the alarm logic flow, and configuring component parameters to meet actual needs. The model editing and testing stages provide functions for saving progress and testing models, ensuring the correctness of model logic and the accuracy of alarm triggering conditions. Finally, in the model management stage, users can easily manage and optimize the model library to achieve real-time monitoring and early warning of the pump station group's operating status.

[0015] This method reduces the difficulty and complexity of model development through visual model building and testing functions, and improves the efficiency and accuracy of alarm diagnosis. At the same time, the model management function allows users to easily manage and optimize the model library, further enhancing the intelligence level of comprehensive alarm diagnosis for pump station groups. In addition, this method also provides real-time monitoring and early warning functions, which can promptly detect and handle potential alarm risks, ensuring the safe and stable operation of pump station groups.

[0016] These model types include threshold alarm, rate alarm, dead point alarm, jitter alarm, and multi-point deviation alarm. Threshold alarm: When the monitored measurement value exceeds the preset upper and lower limits, a threshold alarm will be triggered. This alarm method is simple and direct and can quickly identify abnormal values ​​that exceed the normal range. Rate alarm: By monitoring the rate of data change, a rate alarm will be triggered when the rate of change exceeds a set threshold. This method can capture the trend of data change and detect potential anomalies in a timely manner. Dead spot alarm: The dead spot alarm focuses on the situation where the data hardly changes within a set time window. When the data remains constant or changes very little for a long time, it may mean that the equipment is malfunctioning or the sensor is not working. At this time, the dead spot alarm will be triggered. Jitter alarm: The jitter alarm focuses on situations where data fluctuates frequently within a short period of time. When the data fluctuation exceeds the normal range, it may mean that the device is unstable or subject to external interference, which will trigger the jitter alarm. Multi-point deviation alarm: By comparing the deviation of a single measuring point from the overall level of multiple measuring points, a multi-point deviation alarm will be triggered when the deviation exceeds a set threshold. This method can identify local anomalies and improve the accuracy of the alarm.

[0017] The warning model includes time-limited warning, residual warning, and probability warning; Time Limit Warning: Monitors whether a task or process is completed within the specified time. If a task or process is not completed within the time limit, a time limit warning will be triggered to remind relevant personnel to take timely measures. Residual warning: Based on the prediction model, when the residual between the actual value and the model prediction value exceeds the set threshold, a residual warning will be triggered. This method can predict future abnormal trends and take measures in advance to avoid failures. Probability warning: The probability of an event is calculated by a probability model. When the probability exceeds a set threshold, a probability warning will be triggered. This method can quantify risk and provide a scientific basis for decision-making. Alarm modeling allows users to set alarm trigger conditions through drag-and-drop operations. Alarm settings enable free configuration and advanced calculations of alarm points. Processed alarm points can be used as the basis for judging other alarms, allowing for further combined use. The platform supports various functions and logical operations, including basic functions for finding maximum, minimum, average, summation, and difference, as well as advanced functions for square root, modulo, truncation, and offset.

[0018] The platform has built-in multiple alarm algorithm models that can be directly applied. Supported alarm algorithm models include threshold alarm, rate alarm, dead point alarm, jitter alarm, and deviation alarm. In addition, the alarm models also include a variety of early warning models, such as time limit early warning, residual early warning, and probability early warning, to meet the early warning needs of different scenarios.

[0019] Practical application: Architecture design, the overall architecture includes data acquisition, data analysis, human-computer interaction and data storage.

[0020] (1) Data acquisition mainly involves collecting data from various devices and sensors; (2) Data analysis mainly generates relevant information that can be used for alarms, including core algorithms for alarm modeling, hierarchical classification and diagnostic analysis; (3) The human-computer interaction is mainly to enable users to easily configure alarm models, view alarm information and manage alarm records; (4) Data storage mainly includes important information such as system configuration, alarm records, and historical data to ensure data integrity and traceability; Interface management allows you to add commonly used network communication devices that support MODBUS TCP and OPC, and configure the device's communication parameters, such as IP address and port number, to ensure that the device can connect and communicate normally. It can also be done through API interface and MQ message queue. Measurement point management involves establishing devices and measurement points. Measurement points are divided into actual device measurement points and memory variable measurement points. The creation of actual device measurement points requires adding the device and confirming its ability to connect and communicate normally. Measurement point information includes the measurement point name, data type, measurement range, and alarm threshold. Memory variable measurement points can be created directly without binding to a device. Furthermore, this page provides editing and deletion functions for measurement points to facilitate effective maintenance and management of measurement information. System Functions 1. Alarm monitoring, (1) Monitoring screen, including the monitoring homepage, real-time alarm page, historical alarm page and alarm diagnosis and analysis page, 1) Centralized Control and Monitoring Homepage The left side of the control center homepage uses images of pump stations as a background to form category modules. Each module displays the current number and level of alarms for the corresponding pump station. The flashing colors visually indicate the current alarm level at each pump station; for example, red flashing represents a Level 1 alarm, and yellow flashing represents a Level 2 alarm. Clicking on each pump station module will take you to the pump station monitoring homepage. The upper half of the right side displays a real-time alarm list, showing all current alarms from all pump stations. The lower half allows switching between alarm statistics bar charts and alarm details. Clicking on an alarm in the list displays the corresponding alarm details: 1. Alarm Monitoring Panel.

[0021] The monitoring screen mainly includes the monitoring homepage, real-time alarm page, historical alarm page, and alarm diagnosis and analysis page.

[0022] 2) Pump station monitoring homepage Similar to the layout of the central control center's homepage, the left side of the pump station monitoring homepage is categorized by main unit, auxiliary equipment, utility equipment, and power distribution equipment. Each category contains corresponding devices, and these device modules record the current number and severity level of alarms. Clicking on any device will take you to the corresponding real-time alarm page. The upper half of the right side displays the current pump station's real-time alarm list, while the lower half displays alarm details. Clicking on an alarm in the list will display its corresponding alarm details. 2. Real-time alarm The real-time alarm monitoring page displays all current devices in the pump station on the left side of the device tree, allowing for single or multiple device selection. The right side displays the real-time alarms of the currently selected device and generates a radar chart, which scores the current device based on a corresponding algorithm. The right side also lists all current real-time alarm information in chronological order of the selected device and the alarm's occurrence time. Each alarm includes the alarm device name, alarm parameters, alarm level, and alarm time. Each alarm is followed by action options, allowing users to categorize the alarm into several states: confirmed, not resolved, and resolved.

[0023] Real-time alarms are listed in a list format, ordered by the selected device and the time the alarm occurred. Each alarm message includes the alarm device name, alarm parameters, alarm level, and alarm time. Operators can handle each alarm. The historical alarm display is similar to the real-time alarm monitoring page. The device tree on the left displays all the current devices in the pump station, and you can select one or more devices. The right side displays the historical alarms of the currently selected device.

[0024] To facilitate quick retrieval of specific alarms, an alarm filtering function is provided, allowing users to filter alarms by site, alarm level, alarm device type, and alarm time range. Additionally, a search box is offered, allowing users to enter keywords (such as device name and alarm parameters) to perform fuzzy matching searches for relevant alarm information.

[0025] Diagnostic analysis: Alarm diagnostic analysis displays the current value, alarm level, alarm category, alarm time, alarm content, alarm status, diagnostic cause and suggested measures of the alarm. It can also generate the value change trend of relevant measuring points that triggered the current alarm a period of time ago. Access control, specifically user access control, is used to manage the access and operation permissions of different users, ensuring system security and data confidentiality. This page provides two main functions: role management and user management.

[0026] Role management allows users to add, edit, and delete roles, and assign corresponding permissions to each role. Permissions include access to specific pages, viewing alarm information, configuring alarm models, and managing measurement points. By assigning different roles to different users, the system's access and operational scope can be flexibly controlled, improving system security.

[0027] User management allows users to add, edit, and delete users, and assign roles to them. User information includes username, password, name, and department. Administrators can view all user information through the user management page and modify or delete it as needed. Administrators can also reset user passwords to ensure user account security.

[0028] Based on the organizational structure of pump station operation and maintenance management, the roles are divided into operators, maintenance personnel, and administrators, and corresponding operating permissions are assigned to each role.

[0029] (1) Operators: Primarily responsible for daily equipment monitoring, viewing and processing alarm information, as well as recording and feedback of alarm information. They can view real-time alarms and historical alarms, but do not have the authority to set alarms, modify alarm rules, or modify and delete measurement points.

[0030] (2) Maintenance personnel: Responsible for handling unresolved alarms and setting and optimizing alarm rules. They can view and modify alarm rules, conduct in-depth analysis of alarm information, and perform equipment maintenance and repair based on the analysis results.

[0031] (3) Administrators: They have the highest authority and can add, delete, and modify user roles, manage all equipment and measurement points, and review and publish alarm rules. They ensure the safe and stable operation of the system and prevent unauthorized access and operation.

[0032] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent alarm diagnosis and smart monitoring of pump stations, characterized in that, Specifically, the following steps are included: Step 1: Preliminary alarm diagnosis. In the preliminary alarm diagnosis stage, the built-in algorithm model is mainly used for alarm diagnosis. Step 2: Early warning analysis. Based on the initial alarm diagnosis, further early warning analysis is conducted using the early warning model. Step 3: Model Building. In the model building stage, select the required components from the component library by dragging and dropping, and connect the components according to the alarm logic flow to build a visual alarm diagnosis model. The connections between components represent data flow or logical dependencies, making the model structure clear and easy to understand. At the same time, configure component parameters such as alarm threshold, alarm level and classification and alarm information to meet actual needs. Step 4: Model Editing and Testing. During the model editing process, you can click the save button at any time to save the current progress and ensure that the work is not lost due to unexpected interruptions. At the same time, a model testing function is provided. Click the test button to run the current model to verify the correctness of the logic and whether the alarm triggering conditions meet the expectations. The test results will be displayed at the bottom of the page, including the triggering status, intermediate calculation results and detailed information of the final output, which is convenient for adjusting and optimizing the model. Step 5: Model Management. The model management page displays all created models and supports editing, copying, enabling, disabling, and deleting models. Enabled models will be automatically applied to the real-time data stream for alarm diagnosis and analysis, enabling real-time monitoring and early warning of the pump station group's operating status. Through the model management function, the model library can be easily managed and optimized, improving the efficiency and accuracy of alarm diagnosis.

2. The method for intelligent alarm diagnosis and smart monitoring of pump stations according to claim 1, characterized in that: The model types in step one include threshold alarm, rate alarm, dead spot alarm, jitter alarm, and multi-point deviation alarm. Threshold alarm: When the monitored measurement value exceeds the preset upper and lower limits, a threshold alarm will be triggered. This alarm method is simple and direct and can quickly identify abnormal values ​​that exceed the normal range. Rate alarm: By monitoring the rate of data change, a rate alarm will be triggered when the rate of change exceeds a set threshold. This method can capture the trend of data change and detect potential anomalies in a timely manner. Dead spot alarm: The dead spot alarm focuses on the situation where the data hardly changes within a set time window. When the data remains constant or changes very little for a long time, it may mean that the equipment is malfunctioning or the sensor is not working. At this time, the dead spot alarm will be triggered. Jitter alarm: The jitter alarm focuses on situations where data fluctuates frequently within a short period of time. When the data fluctuation exceeds the normal range, it may mean that the device is unstable or subject to external interference, which will trigger the jitter alarm. Multi-point deviation alarm: By comparing the deviation of a single measuring point from the overall level of multiple measuring points, a multi-point deviation alarm will be triggered when the deviation exceeds a set threshold. This method can identify local anomalies and improve the accuracy of the alarm.

3. The method for intelligent alarm diagnosis and smart monitoring of pump stations according to claim 1, characterized in that: The early warning model in step two includes time-limited early warning, residual early warning, and probability early warning; Time Limit Warning: Monitors whether a task or process is completed within the specified time. If a task or process is not completed within the time limit, a time limit warning will be triggered to remind relevant personnel to take timely measures. Residual warning: Based on the prediction model, when the residual between the actual value and the model prediction value exceeds the set threshold, a residual warning will be triggered. This method can predict future abnormal trends and take measures in advance to avoid failures. Probability warning: By calculating the probability of an event occurring through a probability model, a probability warning will be triggered when the probability exceeds a set threshold. This method can quantify risk and provide a scientific basis for decision-making.

Images