Energy abnormality alarm pre-closed loop management method and system based on hierarchical response

Abstract

The application provides a hierarchical response-based energy abnormality alarm pre-closed loop management method and system, belonging to the technical field of shipbuilding energy management. The method adopts a six-layer modular architecture of perception layer, network layer, data layer, analysis layer, business layer and display layer, and collects multi-source energy data such as power, gas, steam, compressed air and water in real time; through intelligent algorithm, four types of energy abnormality events, including edge device abnormality, energy consumption abnormality, power failure abnormality and deviation rate abnormality, are identified; based on three dimensions of emergency degree, influence range and disposal difficulty, a dynamic weight adjustment algorithm is used to divide the alarm into five levels, match three intelligent push combinations of platform, short message and telephone and corresponding response strategies; through four core links of alarm triggering and confirmation, hierarchical response and disposal, result verification and closed loop, analysis improvement and prevention, full-process pre-closed loop management is realized. The response efficiency and disposal standardization of energy abnormality alarm of shipbuilding enterprises are significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of energy management technology in shipbuilding, and more specifically, relates to a pre-closed-loop control method and system for energy anomaly alarm based on graded response. Background Technology

[0002] Shipbuilding is a typical high-energy-consuming, multi-process, and large-scale industry. Its energy management system covers various energy media such as electricity, natural gas, compressed air, steam, and water, and involves multiple key production areas, including large shipyards, pre-treatment production lines, painting workshops, and section manufacturing workshops. The equipment is diverse and widely distributed. During the energy production, transmission, and use process, abnormal events such as equipment failure, excessive energy consumption, power outages, and parameter deviations occur frequently. If these events are not detected and handled in a timely manner, they will directly affect production safety, cause energy waste, and result in significant economic losses.

[0003] Currently, the energy anomaly alarm handling methods commonly used by shipbuilding companies have the following significant drawbacks: The alarm classification and response mechanism is unreasonable: existing systems mostly adopt simple threshold alarm mechanisms without scientifically classifying abnormal events according to their urgency, scope of impact, and difficulty of handling. This results in a large number of low-priority alarms overwhelming critical alarms. At the same time, the uniform response strategy does not distinguish between alarm levels and types, and high-level emergency alarms are treated together with low-level ordinary alarms, which seriously delays the opportunity to deal with critical issues.

[0004] Alarm notification methods are limited and inflexible: Traditional systems rely solely on platform pop-up notifications and lack multiple notification channels such as SMS and telephone. They cannot dynamically select the optimal notification method based on the alarm level and the status of on-duty personnel. Especially during off-duty hours or in emergencies, information transmission efficiency is low, and important alarms may be missed due to poor notification channels.

[0005] The handling process is not standardized and lacks a closed loop: The existing system only provides alarm information display and does not have a pre-set standardized handling process guide. Different maintenance personnel have different handling methods, and the handling quality is difficult to guarantee. At the same time, there is a lack of a complete alarm handling tracking, feedback and verification mechanism, and the handling results cannot be traced. There are situations where the handling is not in place and the problem is repeated, making it difficult to form an effective management closed loop.

[0006] Multi-source alarms are scattered and lack correlation analysis: various alarms such as edge device alarms, energy consumption alarms, power outage analysis alarms, and deviation rate alarms are managed separately, lacking a unified monitoring interface and cross-type correlation analysis capabilities. Operation and maintenance personnel cannot fully grasp the overall operating status of the energy system and find it difficult to quickly locate the root cause of anomalies.

[0007] With the deepening of digital transformation in shipbuilding enterprises, the existing methods for handling energy anomaly alarms can no longer meet the needs of intelligent and refined management. There is an urgent need for an energy anomaly alarm control solution that can achieve accurate identification, graded response, standardized handling, and closed-loop management. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention provides a pre-closed-loop control method and system for energy anomaly alarms based on hierarchical response. This system solves the technical problems in existing shipbuilding energy management systems, such as unreasonable alarm hierarchies, single notification methods, non-standard handling procedures, lack of correlation among multi-source alarms, and absence of complete closed-loop management. These problems lead to key anomalies being overwhelmed, untimely responses, inconsistent handling quality, and recurring similar issues.

[0009] The energy anomaly alarm pre-closed-loop control method based on hierarchical response includes the following steps: S1 Multi-Source Energy Data Acquisition: Through IoT sensing devices deployed at the shipbuilding site, energy data of electricity, gas, steam, compressed air, and water are collected in real time and transmitted to a unified energy data center for storage via industrial Ethernet, 4G / 5G, or WiFi networks. The transmission process supports breakpoint resume and data caching. S2 Multi-type Energy Anomaly Detection: Real-time analysis of collected energy data to identify four types of energy anomaly events: edge device anomaly, energy consumption exceeding the standard anomaly, power outage anomaly, and deviation rate anomaly. S3 Multi-Dimensional Dynamic Grading Judgment: Based on a preset five-level grading system, combined with three core dimensions of the urgency, scope of impact, and difficulty of handling of abnormal events, a dynamic weight adjustment algorithm is used to calculate the comprehensive risk value of abnormal events, and the alarm level is determined based on the comprehensive risk value; S4 Tiered Intelligent Response Push: Based on the determined alarm level, it automatically matches the corresponding response strategy, selects the optimal push combination from three push methods (platform push, SMS push, and telephone push), pushes the alarm information to the corresponding responsible personnel, and automatically escalates alarms that are not confirmed within a specified time. S5 Full-Process Closed-Loop Handling: After receiving an alarm, the responsible personnel complete the abnormal handling according to the standardized handling process. The system automatically or manually verifies the handling result. If the verification is successful, the alarm is closed; if the verification fails, the alarm is retried to enter a new round of handling. S6 Continuous Analysis and Optimization: After alarm closure, an alarm analysis report is generated. Root cause analysis is used to identify the underlying causes of anomalies, optimize alarm rules, threshold settings, and handling procedures, and achieve a shift from passive alarm to proactive prevention.

[0010] Preferably, the detection method for the four types of abnormal energy events in step S2 is as follows: Edge device anomaly detection: A heartbeat detection mechanism is used to monitor the online status of intelligent sensing devices in real time. When communication interruption, abnormal data acquisition, device offline or sensor drift is detected, an edge device alarm is triggered, and automatic device reconnection and fault self-recovery are supported. Energy consumption exceeding the standard detection: The actual energy consumption of each process is compared with the preset energy consumption standards at the process level, production line level, workshop level, and enterprise level. When the actual energy consumption exceeds the threshold, an energy consumption exceeding the standard alarm is triggered, and trend prediction is performed in combination with historical data to achieve proactive early warning. Power outage anomaly detection: Real-time monitoring of voltage, frequency, and harmonic parameters of the power system. When a power outage, voltage drop, frequency fluctuation, or harmonic anomaly is detected, a power outage alarm is triggered, and the outage time, duration, and affected area are recorded. Deviation rate anomaly detection: Based on historical data, establish a benchmark model for energy consumption of each process. When the deviation rate between the actual energy consumption and the benchmark energy consumption exceeds the allowable range, a deviation rate alarm is triggered.

[0011] Preferably, the five-level hierarchical system and corresponding response strategy described in step S3 are as follows: The highest level red alarm corresponds to a comprehensive risk value of ≥90 points. It is applicable to large-scale power outages, critical equipment failures, and major energy consumption anomalies. It requires intervention from senior management personnel of the enterprise and uses a triple push method of telephone + SMS + platform. Advanced Orange Alarm: Corresponds to a comprehensive risk value of 70-89 points. It is applicable to workshop-level energy consumption exceeding standards, abnormality of important equipment, and large-scale power quality problems. It is to be responded to by workshop supervisors and above, and is pushed by both telephone and SMS. Intermediate Yellow Alarm: Corresponds to a comprehensive risk value of 50-69 points. It is applicable to abnormal energy consumption at the production line level, general equipment problems, and local power fluctuation events. It is to be responded to by the team leader or above and is delivered via SMS and platform push. Low-level blue alarm: Corresponds to a comprehensive risk value of 30-49 points. It is applicable to single-point device alarms, small energy consumption fluctuations, and minor deviation events. It is handled by maintenance personnel during normal working hours and is pushed through the platform. The lowest level green alarm corresponds to a comprehensive risk value of <30 points. It is applicable to edge device heartbeat abnormalities and minor abnormalities in data acquisition delays. It is only recorded by the system and handled uniformly during routine inspections.

[0012] Preferably, the three push methods described in step S4 are implemented as follows: Platform push notifications: Alarm details, real-time data, trend charts, and handling suggestions are pushed via real-time pop-ups on the web management terminal and mobile APP; SMS push: It generates alarm SMS messages using templated and personalized content, supports scheduled resending and unconfirmed escalation mechanisms, and is suitable for medium and high-level alarms; Telephone push: It adopts a combination of voice broadcast and manual confirmation, and is only used for the highest level and advanced alarms. It supports automatic dialing, voice broadcast, reception confirmation and escalation transfer, and can be connected with the enterprise emergency command system to realize one-click calling.

[0013] Preferably, the closed-loop processing described in step S5 includes four core steps: Alarm Triggering and Confirmation: The system automatically performs intelligent filtering and deduplication to exclude duplicate and invalid alarms and pushes alarm information to the responsible personnel. If the alarm is not confirmed within the specified time, the push level will be automatically escalated. Tiered Response and Handling: Based on the alarm level, a corresponding standardized handling procedure template is matched, and the responsible personnel record the handling process and results in real time; Result verification and closed-loop: For anomalies that can be automatically detected and recovered, the system monitors relevant data indicators in real time to verify the recovery status; for anomalies that require manual confirmation, the responsible personnel are required to upload on-site photos, measurement data or sign for confirmation. If the verification is successful, the alarm is closed-loop; otherwise, the alarm is retried. Quantitative evaluation of handling effectiveness: The system quantifies the handling effectiveness based on three indicators: handling time, degree of anomaly recovery, and repeat alarm rate. The scoring results are included in the performance evaluation of the responsible personnel.

[0014] Preferably, the method is implemented based on a six-layer system architecture, wherein the six-layer system architecture is as follows: The sensing layer includes IoT sensing devices such as smart meters, gas sensors, temperature sensors, and flow meters, and supports multiple communication protocols and data formats. Network layer: Employs industrial Ethernet, 4G / 5G mobile communication, and WiFi network technologies to achieve reliable transmission and real-time aggregation of data from the perception layer; Data layer: Construct a unified energy data center, using a time-series database to store real-time data, a relational database to store configuration data, and distributed storage to store historical data; Analysis layer: Utilizing big data analytics, machine learning, and edge computing technologies, it performs real-time analysis, anomaly detection, trend prediction, and correlation analysis on energy data; Business layer: Implements core business functions such as alarm management, hierarchical response, closed-loop handling, and statistical analysis, and supports flexible process configuration, permission management, and role allocation; Presentation layer: Provides multiple display methods including web management terminal, mobile APP, and large screen monitoring, and supports real-time monitoring screen, alarm list, statistical analysis report, and GIS map visualization functions.

[0015] Another technical problem to be solved by this invention is to provide a pre-closed-loop control system for energy anomaly alarms based on hierarchical response, comprising: The multi-source data acquisition module is used to collect energy data of electricity, gas, steam, compressed air and water in real time through IoT sensing devices deployed at the shipbuilding site, and transmit them to the data storage module. The data storage module is used to build a unified energy data center. It adopts a combination of time-series databases, relational databases and distributed storage to realize the unified storage, management and retrieval of multi-source heterogeneous energy data. The multi-type anomaly detection module is used to perform real-time analysis of the collected energy data and identify four types of energy anomaly events: edge device anomaly, energy consumption exceeding the standard anomaly, power outage anomaly, and deviation rate anomaly. The multi-dimensional dynamic grading module is used to calculate the comprehensive risk value of abnormal events based on a preset five-level grading system, combined with three core dimensions: the urgency of the abnormal event, the scope of its impact, and the difficulty of handling it, and to determine the alarm level by using a dynamic weight adjustment algorithm. The hierarchical intelligent push module is used to automatically match the corresponding response strategy according to the alarm level, select the optimal push combination from three push methods: platform push, SMS push, and telephone push, push alarm information to the corresponding responsible personnel, and automatically escalate the push for alarms that are not confirmed in time. The closed-loop handling module supports responsible personnel in handling anomalies according to standardized procedures, automatically or manually verifying the handling results, and realizing full lifecycle management of alarm events. The continuous optimization module is used to generate an alarm analysis report after the alarm loop is closed. It identifies the underlying causes of anomalies through root cause analysis and optimizes alarm rules, threshold settings, and handling procedures.

[0016] Preferably, the hierarchical intelligent push module includes: The platform push unit is used to push alarm details, real-time data, trend charts, and handling suggestions through real-time pop-ups on the web management terminal and mobile APP. The SMS push unit is used to generate templated and personalized alarm SMS messages, and supports scheduled resending and unconfirmed escalation mechanisms; The telephone push unit is used to realize voice broadcast and manual confirmation functions, supports automatic dialing, receiving confirmation and escalation transfer, and can be connected to the enterprise emergency command system; The push strategy dynamic configuration unit is used to dynamically adjust the push method and the priority of the push targets based on the alarm level, the on-duty status of the staff, and the historical response efficiency.

[0017] Preferably, the closed-loop processing module includes: The alarm filtering and confirmation unit is used to automatically filter duplicate alarms and invalid alarms, track the alarm confirmation status, and automatically escalate alarms that are not confirmed in a timely manner. The standardized handling guidance unit is used to store standardized handling process templates for different types and levels of alarms, providing step-by-step operation guidance for responsible personnel. The result verification unit is used to automatically monitor the recovery status of detectable anomalies and supports manual uploading of disposal evidence for verification. The quantitative evaluation unit for handling effectiveness is used to quantitatively score the handling effectiveness based on three indicators: handling time, degree of abnormal recovery, and recurrence alarm rate, and generate a handling performance statistical report.

[0018] Preferably, it also includes a visualization module, which provides three display terminals: a web management terminal, a mobile APP, and a large-screen monitoring terminal, and supports: The centralized display and cross-type correlation analysis of four types of alarms enable unified monitoring of the overall operating status of the energy system. Multi-dimensional statistical analysis reports showcasing alarm time distribution, type distribution, response time, and repeat alarm rate; Energy pipeline network GIS map visualization, intuitively showing the location and impact range of anomalies; Real-time tracking of disposal progress and full-process traceability and query of historical disposal records.

[0019] Compared with the prior art, the present invention has the following beneficial effects: Significantly improve alarm response efficiency and ensure priority handling of critical anomalies: Through a scientific five-level classification system and a multi-dimensional dynamic weight adjustment algorithm, differentiated alarm processing and optimized allocation of operation and maintenance resources are achieved. This enables operation and maintenance personnel to quickly identify and prioritize high-level alarms, avoiding interference from a large number of low-level alarms, resulting in an overall response efficiency improvement of over 60%. An intelligent combination mechanism of platform, SMS, and telephone push notifications, combined with an automatic escalation function for unconfirmed alerts, ensures that alarm information reaches the responsible personnel immediately during non-working hours or in emergency situations, achieving a 100% reach rate for important alarms.

[0020] Standardized handling processes enable closed-loop management throughout the entire lifecycle: Standardized handling process templates are pre-set for four typical anomaly scenarios, providing step-by-step operation guidance for maintenance personnel, unifying handling standards, reducing reliance on personnel experience, and improving handling standardization by over 80%. Through the close connection of four stages—alarm triggering and confirmation, tiered response and handling, result verification and closure, and analysis, improvement, and prevention—full lifecycle management of alarm events is achieved, ensuring that every alarm is effectively handled, the problem is thoroughly resolved, and the recurrence rate is reduced by over 50%. A new quantitative evaluation function for handling effectiveness has been added, scoring handling effectiveness from three dimensions: handling time, anomaly recovery degree, and recurrence rate. The scoring results are incorporated into the performance evaluation of responsible personnel, further improving the quality of handling.

[0021] Achieving unified monitoring of multi-source alarms and improving the accuracy of anomaly identification: A unified alarm monitoring dashboard is constructed to centrally display four types of alarms and perform cross-type correlation analysis. Maintenance personnel can comprehensively grasp the overall operating status of the energy system and quickly locate the root cause of anomalies. Through the organic combination of four functions—edge device alarms, energy consumption exceeding limits alarms, power outage alarms, and deviation rate alarms—full coverage and accurate identification of various energy anomalies during shipbuilding are achieved, effectively preventing important anomalies from being missed or overlooked.

[0022] Enhancing system intelligence and operational reliability: Adopting a six-layer modular system architecture, each layer functions independently yet collaboratively, supporting elastic expansion and customized deployment to meet the energy management needs of shipbuilding enterprises of different sizes. System stability and maintainability are significantly improved. Utilizing big data analytics, machine learning, and cloud-edge collaboration technologies, proactive early warning and root cause analysis of energy anomalies are achieved, driving the transformation of energy management from passive alarm to proactive prevention.

[0023] Reduce operation and maintenance costs and promote continuous optimization of energy management: Intelligent management reduces the workload of manual inspections and notifications, standardized processes reduce personnel training costs, and closed-loop management reduces losses caused by recurring problems, resulting in an overall reduction of operation and maintenance costs by more than 30%. The system automatically generates multi-dimensional alarm analysis reports, providing data support and decision-making basis for enterprise energy management optimization, promoting continuous improvement in energy utilization efficiency, and helping enterprises achieve energy conservation and consumption reduction goals. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the alarm recording interface of the edge device of the present invention; Figure 2 This is a schematic diagram of the energy consumption over-limit alarm interface of the present invention; Figure 3 This is a schematic diagram of the power outage analysis and alarm interface of the present invention; Figure 4 This is a schematic diagram of the deviation rate alarm recording interface of the present invention; Figure 5 This is a schematic diagram of the edge alarm push configuration list interface of the present invention; Figure 6 This is a schematic diagram of the newly created device push configuration interface of the present invention; Figure 7 This is a schematic diagram of the alarm push rule details interface of the present invention; Figure 8 This is a schematic diagram of the newly created server push configuration interface of the present invention; Figure 9 This is a schematic diagram of the pre-closed-loop handling method for energy anomaly alarms of the present invention; Figure 10 This is a schematic diagram of the six-layer architecture of the energy anomaly alarm pre-closed-loop control system of the present invention. Detailed Implementation

[0025] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0026] This invention discloses a pre-closed-loop control method and system for energy anomaly alarms based on hierarchical response. It is applicable to energy monitoring and anomaly handling scenarios in key shipbuilding processes such as large shipyards, pre-treatment production lines, painting workshops, and section manufacturing workshops. Through a six-layer modular system architecture, four types of full-coverage alarm mechanisms, a five-level dynamic hierarchical system, three intelligent push methods, and a four-stage closed-loop management process, it solves the technical problems in existing technologies such as unreasonable alarm hierarchies, single notification methods, non-standard handling processes, and lack of closed-loop management.

[0027] I. System Overall Architecture: This invention adopts a six-layer modular system architecture, where each layer functions independently yet works collaboratively, supporting elastic expansion and customized deployment: 1. Perception layer: Deployed across all energy nodes in the shipbuilding plant, these devices include IoT sensing equipment such as smart meters, smart gas meters, steam flow meters, compressed air pressure sensors, temperature sensors, water quality sensors, and equipment status acquisition terminals. All devices support mainstream industrial communication protocols such as Modbus, OPCUA, and MQTT, and the sampling frequency is configurable (1 second / time for power data, 5 seconds / time for water and gas data).

[0028] The edge device has a built-in heartbeat detection module that sends a heartbeat packet to the system every 30 seconds. If no heartbeat is received for three consecutive times, the device is considered offline. It supports local data caching, which can cache data for 72 hours when the network is interrupted and automatically resume transmission after the network is restored, ensuring data integrity.

[0029] 2. Network layer: Construct a hybrid transmission network with "industrial Ethernet as the primary mode, 4G / 5G as a secondary mode, and WiFi as a supplement": Gigabit industrial Ethernet is used in fixed areas such as the workshop, substation, and energy station to ensure the real-time performance and reliability of data transmission. 5G private networks are used in mobile operation areas such as docks and wharves to support high-bandwidth, low-latency data transmission; Temporary work sites utilize WiFi 6 wireless access for rapid deployment and flexible expansion. A firewall and intrusion detection system are configured at the network layer to achieve physical isolation between the industrial and office networks, ensuring data security.

[0030] 3. Data Layer: Build a unified energy data center and adopt a hybrid storage architecture: InfluxDB (Time Series Database): Stores real-time energy data and device status data, supports millions of data writes per second, and has a retention period of 3 years; Relational databases (MySQL): Store structured data such as system configuration data, user information, alarm rules, and handling procedures; The Distributed File System (HDFS) stores unstructured data such as historical data, alarm records, and evidence of handling (photos, videos, documents), with a retention period of 10 years. The data layer provides a unified data access interface, supporting SQL queries and API calls to achieve unified management of multi-source heterogeneous data.

[0031] 4. Analysis layer: Deployed on cloud platforms and edge nodes, it adopts a "cloud-edge collaborative" analytics architecture: Edge nodes: perform preprocessing and initial anomaly detection on real-time data, filter invalid data, and reduce cloud computing pressure; Cloud-based analytics platform: Integrates a big data analytics engine and machine learning framework, providing core algorithms such as anomaly detection, trend prediction, correlation analysis, and root cause analysis. The analytics layer supports custom alarm rules and analysis models, allowing users to configure thresholds, trigger conditions, and analysis cycles according to actual production needs.

[0032] 5. Business layer: Implement the system's core business functions, including: Alarm Management Module: Responsible for alarm triggering, filtering, deduplication, classification, and push notifications; Tiered response module: Matches corresponding response strategies and handling procedures based on alarm levels; Closed-loop handling module: Tracks the entire alarm handling process, enabling result verification and closed-loop management; Statistical analysis module: Generates multi-dimensional statistical reports and analysis reports; System management module: responsible for functions such as user permissions, role assignment, process configuration, and data backup.

[0033] 6. Presentation Layer: Three display terminals are provided: Web management interface: Designed for administrators and operations personnel, providing a complete system function operation interface; Mobile App: Designed for on-site personnel, supporting functions such as alarm reception, incident reporting, and result verification; Large-screen monitoring: Deployed in the energy management center, it displays the operating status of the entire plant's energy system, real-time alarm information, and statistical data. The display layer supports GIS map visualization, which can intuitively show the energy pipeline topology, equipment location, and the location of anomalies.

[0034] II. Core Functional Modules: 1. Four types of energy anomaly alarm functions: (1) Edge device alarm: Perform full lifecycle status monitoring on all sensing devices, with triggering conditions including: Communication interruption: No heartbeat packets were received from the device for three consecutive times; Data acquisition anomaly: The acquired data exceeds the device's measurement range or shows fixed or fluctuating values; Device offline: The device is powered off or the network connection is interrupted; Sensor drift: The deviation of the collected data from the baseline value exceeds 5% after 24 consecutive hours. After the alarm is triggered, the system automatically records information such as the device number, the type of abnormality, and the time of occurrence, and attempts to automatically reconnect to the device. If the reconnection is successful, the alarm will be automatically restored.

[0035] (2) Energy consumption exceeding the standard alarm: It supports four levels of energy consumption index configuration: process level, production line level, workshop level, and enterprise level. The trigger condition is: actual energy consumption > standard energy consumption × (1 + allowable deviation rate). The standard energy consumption can be dynamically adjusted according to production plans, product models, seasonal factors, etc. The system uses the ARIMA algorithm to predict energy consumption trends based on historical data. When it is predicted that energy consumption will exceed the standard in the next 24 hours, an early warning message will be issued.

[0036] (3) Power outage analysis alarm: Real-time monitoring of power system parameters such as voltage, current, frequency, and harmonics, triggered by the following conditions: Power outage: All three phase voltages are below 10% of the rated voltage and the duration exceeds 1 second; Voltage drop: The voltage drops below 70% of the rated voltage and lasts for more than 0.5 seconds; Frequency fluctuation: The frequency exceeds the range of 49.5-50.5Hz and lasts for more than 2 seconds; Harmonic anomaly: Total harmonic distortion rate exceeds 5%. After the alarm is triggered, the system automatically records the power outage start time, affected area, affected equipment, and other information, and generates a power outage event report.

[0037] (4) Deviation rate analysis alarm: Based on historical data, a benchmark model for energy consumption in each process is established. The deviation rate is calculated using the formula: Deviation Rate = |Actual Energy Consumption - Benchmark Energy Consumption| / Benchmark Energy Consumption × 100%. An alarm is triggered when the deviation rate exceeds the preset allowable range (default 2%). The system supports deviation rate calculation for different statistical periods, including hours, days, weeks, and months, and can sensitively detect abnormal fluctuations in energy consumption.

[0038] 2. Five-level dynamic hierarchical system: This invention employs a multi-dimensional dynamic weight adjustment algorithm to classify alarms, comprehensively considering three core dimensions: urgency (weight 40%), impact range (weight 35%), and handling difficulty (weight 25%), to calculate a comprehensive risk value: Comprehensive Risk Value = Urgency Score × 0.4 + Impact Range Score × 0.35 + Handling Difficulty Score × 0.25. The scoring criteria for each dimension are as follows:

[0039] Alarms are classified into five levels based on the overall risk value: Highest level (red): Overall risk score ≥ 90 points, response time limit 5 minutes; Advanced (Orange): Overall risk score 70-89, response time limit 10 minutes; Intermediate (Yellow): Overall risk score 50-69 points, response time limit 30 minutes; Low level (blue): Overall risk score 30-49 points, response time limit 2 hours; Lowest level (green): Overall risk score < 30 points, no immediate response required.

[0040] 3. Three intelligent push notification methods: The system automatically selects the optimal push notification combination based on the alarm level and the status of on-duty personnel: (1) Platform push: The basic push notification method is applicable to all alarm levels. Push notification content includes: alarm number, alarm type, alarm level, occurrence time, location of the anomaly, real-time data, trend charts, and handling suggestions. The web interface uses pop-up windows and sound alerts, while the mobile app uses push notifications and vibration alerts.

[0041] (2) SMS push: This supplementary push notification method is applicable to intermediate and higher-level alarms. The SMS content uses a template design, example: "[Energy Alarm] Highest-level alarm: Main transformer fault at substation #3, affecting the thin plate workshop and painting workshop. Please handle immediately. Alarm time: 2026-04-29 08:23:34". Scheduled resending is supported, once every 5 minutes, up to a maximum of 3 times.

[0042] (3) Telephone push: The highest priority push method is only applicable to the highest and advanced alarms. It uses an automatic voice broadcast + manual confirmation mode. Example voice message: "Hello, this is the Energy Management Center. We are now broadcasting the highest priority alarm: Main transformer fault at substation #3. Affected areas: Thin plate workshop and painting workshop. Please proceed immediately. Please press 1 to confirm receipt." If no confirmation is received within 1 minute, the call will be automatically transferred to the next responsible person, up to a maximum of 3 people.

[0043] Automatic escalation mechanism: If an alarm is not acknowledged within the specified response time limit, the system automatically escalates the push notification level. For example, if a high-level alarm is not acknowledged within 10 minutes, it will be escalated to the highest-level alarm, and a telephone push notification will be added to the system to reach senior management.

[0044] 4. Four-stage closed-loop management process: (1) Alarm triggering and confirmation: The system automatically triggers alarms based on preset rules, first performing intelligent filtering and deduplication: Filtering: Eliminates invalid alarms caused by equipment debugging, planned maintenance, etc. Deduplication: Duplicate alarms of the same type from the same device are merged, retaining only the most recent one. After the alarm information is pushed to the responsible personnel, the system tracks and confirms the status. If confirmation is not received within the specified time, an automatic escalation mechanism is executed.

[0045] (2) Tiered response and handling: The system provides pre-set standardized handling procedure templates for different types and levels of alarms, for example: Highest level power outage alarm: Immediately disconnect non-essential loads → notify emergency power generation unit to start → investigate the cause of the fault → organize emergency repairs → restore power supply; Intermediate energy consumption exceeding limit alarm: View energy consumption details → Locate high-energy-consuming equipment → Check equipment operating status → Adjust operating parameters → Record handling results. Responsible personnel should handle the situation according to the process template, recording the handling steps and intermediate results in the system in real time.

[0046] (3) Result verification and closed loop: After the processing is completed, the system automatically verifies the results: Automatic verification: For anomalies that can be detected and recovered through sensors (such as voltage recovery, energy consumption reduction), the system monitors relevant data in real time, and if the data is normal for 5 consecutive minutes, it is determined to be a recovery; Manual verification: For anomalies that cannot be automatically detected (such as equipment repair or pipeline leak repair), responsible personnel are required to upload on-site photos, measurement data, or signed confirmation as evidence. Upon successful verification, the alarm automatically closes the loop; if verification fails, the system re-triggers the alarm, initiating a new round of handling procedures.

[0047] (4) Analysis, improvement and prevention: After the alarm loop is closed, the system automatically generates an alarm analysis report, which includes: Basic alarm information: time, type, level, location, and personnel handling the alarm; Process record: processing steps, time taken, resources used; Statistical analysis: response time, repeat alarm rate, frequency of similar alarms; Root cause analysis: The 5 Whys analysis method is used to identify the underlying causes of anomalies; Improvement suggestions: Optimize alarm thresholds, adjust equipment maintenance cycles, and improve production processes. The system automatically optimizes alarm rules and handling procedures based on analysis results, achieving continuous improvement.

[0048] III. Complete Method and Flow: The energy anomaly alarm pre-closed-loop control method based on hierarchical response of the present invention comprises the following specific steps: S1 Multi-Source Energy Data Acquisition: IoT sensing devices deployed at the shipbuilding site collect real-time energy data such as electricity, gas, steam, compressed air, and water. The data is transmitted from the network layer to the data layer for unified storage, and the transmission process supports breakpoint resume and local caching.

[0049] S2 Multi-type Energy Anomaly Detection: The analysis layer performs real-time analysis on the collected energy data and identifies four types of energy anomaly events through preset rules and intelligent algorithms: edge device anomaly, energy consumption exceeding the standard anomaly, power outage anomaly, and deviation rate anomaly.

[0050] S3 Multi-Dimensional Dynamic Grading Judgment: The system calculates the comprehensive risk value based on three dimensions of the urgency, scope of impact, and difficulty of handling of the abnormal event, using a dynamic weight adjustment algorithm. Based on the comprehensive risk value, the alarm level is determined, and graded alarm information is generated.

[0051] S4 Tiered Intelligent Response Push: Automatically matches the corresponding response strategy based on the alarm level, selecting the optimal combination of platform push, SMS push, and telephone push to send alarm information to the responsible personnel. If no confirmation is received within the specified time, the push level is automatically escalated.

[0052] S5 Full-Process Closed-Loop Handling: After receiving an alarm, the responsible personnel handle the anomaly according to the standardized handling procedure template, recording the handling process in real time. After the handling is completed, the system confirms whether the anomaly has been eliminated through automatic detection or manual verification. If the verification is successful, the alarm is closed; otherwise, the alarm is retried.

[0053] S6 Continuous Analysis and Optimization: After the alarm loop is closed, the system automatically generates an analysis report, identifies the underlying causes of the anomaly through root cause analysis, proposes improvement suggestions, optimizes alarm rules, threshold settings and handling procedures, and realizes the transformation from passive alarm to proactive prevention. IV. Specific Implementation Examples: Taking the failure of the main transformer in substation No. 3 of the thin plate workshop of a large shipbuilding enterprise as an example, the implementation process of the present invention is described in detail: Data Acquisition and Anomaly Detection: At 08:23:34 on April 29, 2026, the smart meter deployed in Substation No. 3 detected that the voltage of phase A was 0, while the voltages of phases B and C were normal. The analysis layer determined that the power outage of the feeder cabinet circuit was abnormal, triggering a power outage analysis alarm.

[0055] Dynamic risk assessment: The system calculates the comprehensive risk value: urgency score 80 points (may lead to production stoppage in the thin plate workshop), impact range score 80 points (affects both the thin plate workshop and the painting workshop), handling difficulty score 80 points (requires coordination from the electrical maintenance team), comprehensive risk value = 80 × 0.4 + 80 × 0.35 + 80 × 0.25 = 80 points, which is judged as a high-level (orange) alarm.

[0056] Tiered intelligent push notification: The system uses a dual push method of telephone + SMS to send alarm information to the electrical workshop supervisor and on-duty electrician. If no confirmation is received after 10 minutes, the system automatically escalates to the highest level (red) alarm and adds a telephone push to the vice president of production.

[0057] Closed-loop processing: At 08:35, the on-duty electrician confirmed receiving the alarm and went to the scene to handle it; At 08:42, the electrician found that the fuse for phase A of the feeder cabinet was blown and replaced it. At 08:48, power was restored, and the system automatically detected that the voltage had returned to normal. At 08:50, the electrician uploaded on-site photos and maintenance records, the system verified them, and the alarm closed the loop.

[0058] Analysis and Improvement: The system generates an alarm analysis report, and root cause analysis shows that fuse aging is the main cause of the fault. Improvement suggestions include: shortening the substation fuse inspection cycle from 6 months to 3 months, adding temperature sensors to monitor fuse operating status, and automatically updating equipment maintenance plans and optimizing alarm rules.

[0059] This embodiment verifies that the present invention can achieve rapid identification, accurate classification, timely response, and closed-loop handling of energy anomalies, effectively shortening the fault handling time and reducing the risk of production interruption.

[0060] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0061] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A pre-closed-loop control method for energy anomaly alarms based on hierarchical response, characterized in that, Includes the following steps: S1 Multi-Source Energy Data Acquisition: Through IoT sensing devices deployed at the shipbuilding site, energy data of electricity, gas, steam, compressed air, and water are collected in real time and transmitted to a unified energy data center for storage via industrial Ethernet, 4G / 5G, or WiFi networks. The transmission process supports breakpoint resume and data caching. S2 Multi-type Energy Anomaly Detection: Real-time analysis of collected energy data to identify four types of energy anomaly events: edge device anomaly, energy consumption exceeding the standard anomaly, power outage anomaly, and deviation rate anomaly. S3 Multi-Dimensional Dynamic Grading Judgment: Based on a preset five-level grading system, combined with three core dimensions of the urgency, scope of impact, and difficulty of handling of abnormal events, a dynamic weight adjustment algorithm is used to calculate the comprehensive risk value of abnormal events, and the alarm level is determined based on the comprehensive risk value; S4 Tiered Intelligent Response Push: Based on the determined alarm level, it automatically matches the corresponding response strategy, selects the optimal push combination from three push methods (platform push, SMS push, and telephone push), pushes the alarm information to the corresponding responsible personnel, and automatically escalates alarms that are not confirmed within a specified time. S5 Full-Process Closed-Loop Handling: After receiving an alarm, the responsible personnel complete the abnormal handling according to the standardized handling process. The system automatically or manually verifies the handling result. If the verification is successful, the alarm is closed; if the verification fails, the alarm is retried to enter a new round of handling. S6 Continuous Analysis and Optimization: After alarm closure, an alarm analysis report is generated. Root cause analysis is used to identify the underlying causes of anomalies, optimize alarm rules, threshold settings, and handling procedures, and achieve a shift from passive alarm to proactive prevention.

2. The energy anomaly alarm pre-closed-loop control method based on hierarchical response according to claim 1, characterized in that, The detection methods for the four types of abnormal energy events mentioned in step S2 are as follows: Edge device anomaly detection: A heartbeat detection mechanism is used to monitor the online status of intelligent sensing devices in real time. When communication interruption, abnormal data acquisition, device offline or sensor drift is detected, an edge device alarm is triggered, and automatic device reconnection and fault self-recovery are supported. Energy consumption exceeding the standard detection: The actual energy consumption of each process is compared with the preset energy consumption standards at the process level, production line level, workshop level, and enterprise level. When the actual energy consumption exceeds the threshold, an energy consumption exceeding the standard alarm is triggered, and trend prediction is performed in combination with historical data to achieve proactive early warning. Power outage anomaly detection: Real-time monitoring of voltage, frequency, and harmonic parameters of the power system. When a power outage, voltage drop, frequency fluctuation, or harmonic anomaly is detected, a power outage alarm is triggered, and the outage time, duration, and affected area are recorded. Deviation rate anomaly detection: Based on historical data, establish a benchmark model for energy consumption of each process. When the deviation rate between the actual energy consumption and the benchmark energy consumption exceeds the allowable range, a deviation rate alarm is triggered.

3. The energy anomaly alarm pre-closed-loop control method based on hierarchical response according to claim 1, characterized in that, The five-level hierarchical system and corresponding response strategy mentioned in step S3 are as follows: The highest level red alarm corresponds to a comprehensive risk value of ≥90 points. It is applicable to large-scale power outages, critical equipment failures, and major energy consumption anomalies. It requires intervention from senior management personnel of the enterprise and uses a triple push method of telephone + SMS + platform. Advanced Orange Alarm: Corresponds to a comprehensive risk value of 70-89 points. It is applicable to workshop-level energy consumption exceeding standards, abnormality of important equipment, and large-scale power quality problems. It is to be responded to by workshop supervisors and above, and is pushed by both telephone and SMS. Intermediate Yellow Alarm: Corresponds to a comprehensive risk value of 50-69 points. It is applicable to abnormal energy consumption at the production line level, general equipment problems, and local power fluctuation events. It is to be responded to by the team leader or above and is delivered via SMS and platform push. Low-level blue alarm: Corresponds to a comprehensive risk value of 30-49 points. It is applicable to single-point device alarms, small energy consumption fluctuations, and minor deviation events. It is handled by maintenance personnel during normal working hours and is pushed through the platform. The lowest level green alarm corresponds to a comprehensive risk value of <30 points. It is applicable to edge device heartbeat abnormalities and minor abnormalities in data acquisition delays. It is only recorded by the system and handled uniformly during routine inspections.

4. The energy anomaly alarm pre-closed-loop control method based on hierarchical response according to claim 1, characterized in that, The specific implementation of the three push methods mentioned in step S4 is as follows: Platform push notifications: Alarm details, real-time data, trend charts, and handling suggestions are pushed via real-time pop-ups on the web management terminal and mobile APP; SMS push: It generates alarm SMS messages using templated and personalized content, supports scheduled resending and unconfirmed escalation mechanisms, and is suitable for medium and high-level alarms; Telephone push: It adopts a combination of voice broadcast and manual confirmation, and is only used for the highest level and advanced alarms. It supports automatic dialing, voice broadcast, reception confirmation and escalation transfer, and can be connected with the enterprise emergency command system to realize one-click calling.

5. The energy anomaly alarm pre-closed-loop control method based on hierarchical response according to claim 1, characterized in that, The closed-loop processing described in step S5 includes four core steps: Alarm Triggering and Confirmation: The system automatically performs intelligent filtering and deduplication to exclude duplicate and invalid alarms and pushes alarm information to the responsible personnel. If the alarm is not confirmed within the specified time, the push level will be automatically escalated. Tiered Response and Handling: Based on the alarm level, a corresponding standardized handling procedure template is matched, and the responsible personnel record the handling process and results in real time; Result verification and closed-loop: For anomalies that can be automatically detected and recovered, the system monitors relevant data indicators in real time to verify the recovery status; for anomalies that require manual confirmation, the responsible personnel are required to upload on-site photos, measurement data or sign for confirmation. If the verification is successful, the alarm is closed-loop; otherwise, the alarm is retried. Quantitative evaluation of handling effectiveness: The system quantifies the handling effectiveness based on three indicators: handling time, degree of anomaly recovery, and repeat alarm rate. The scoring results are included in the performance evaluation of the responsible personnel.

6. The energy anomaly alarm pre-closed-loop control method based on hierarchical response according to claim 1, characterized in that, The method is implemented based on a six-layer system architecture, which consists of the following layers: The sensing layer includes IoT sensing devices such as smart meters, gas sensors, temperature sensors, and flow meters, and supports multiple communication protocols and data formats. Network layer: Employs industrial Ethernet, 4G / 5G mobile communication, and WiFi network technologies to achieve reliable transmission and real-time aggregation of data from the perception layer; Data layer: Construct a unified energy data center, using a time-series database to store real-time data, a relational database to store configuration data, and distributed storage to store historical data; Analysis layer: Utilizing big data analytics, machine learning, and edge computing technologies, it performs real-time analysis, anomaly detection, trend prediction, and correlation analysis on energy data; Business layer: Implements core business functions such as alarm management, hierarchical response, closed-loop handling, and statistical analysis, and supports flexible process configuration, permission management, and role allocation; Presentation layer: Provides multiple display methods including web management terminal, mobile APP, and large screen monitoring, and supports real-time monitoring screen, alarm list, statistical analysis report, and GIS map visualization functions.

7. A graded response-based energy anomaly alarm pre-closed-loop control system, characterized in that, include: The multi-source data acquisition module is used to collect energy data of electricity, gas, steam, compressed air and water in real time through IoT sensing devices deployed at the shipbuilding site, and transmit them to the data storage module. The data storage module is used to build a unified energy data center. It adopts a combination of time-series databases, relational databases and distributed storage to realize the unified storage, management and retrieval of multi-source heterogeneous energy data. The multi-type anomaly detection module is used to perform real-time analysis of the collected energy data and identify four types of energy anomaly events: edge device anomaly, energy consumption exceeding the standard anomaly, power outage anomaly, and deviation rate anomaly. The multi-dimensional dynamic grading module is used to calculate the comprehensive risk value of abnormal events based on a preset five-level grading system, combined with three core dimensions: the urgency of the abnormal event, the scope of its impact, and the difficulty of handling it, and to determine the alarm level by using a dynamic weight adjustment algorithm. The hierarchical intelligent push module is used to automatically match the corresponding response strategy according to the alarm level, select the optimal push combination from three push methods: platform push, SMS push, and telephone push, push alarm information to the corresponding responsible personnel, and automatically escalate the push for alarms that are not confirmed in time. The closed-loop handling module supports responsible personnel in handling anomalies according to standardized procedures, automatically or manually verifying the handling results, and realizing full lifecycle management of alarm events. The continuous optimization module is used to generate an alarm analysis report after the alarm loop is closed. It identifies the underlying causes of anomalies through root cause analysis and optimizes alarm rules, threshold settings, and handling procedures.

8. The energy anomaly alarm pre-closed-loop control system based on hierarchical response according to claim 7, characterized in that, The hierarchical intelligent push module includes: The platform push unit is used to push alarm details, real-time data, trend charts, and handling suggestions through real-time pop-ups on the web management terminal and mobile APP. The SMS push unit is used to generate templated and personalized alarm SMS messages, and supports scheduled resending and unconfirmed escalation mechanisms; The telephone push unit is used to realize voice broadcast and manual confirmation functions, supports automatic dialing, receiving confirmation and escalation transfer, and can be connected to the enterprise emergency command system; The push strategy dynamic configuration unit is used to dynamically adjust the push method and the priority of the push targets based on the alarm level, the on-duty status of the staff, and the historical response efficiency.

9. The energy anomaly alarm pre-closed-loop control system based on hierarchical response according to claim 7, characterized in that, The closed-loop processing module includes: The alarm filtering and confirmation unit is used to automatically filter duplicate alarms and invalid alarms, track the alarm confirmation status, and automatically escalate alarms that are not confirmed in a timely manner. The standardized handling guidance unit is used to store standardized handling process templates for different types and levels of alarms, providing step-by-step operation guidance for responsible personnel. The result verification unit is used to automatically monitor the recovery status of detectable anomalies and supports manual uploading of disposal evidence for verification. The quantitative evaluation unit for handling effectiveness is used to quantitatively score the handling effectiveness based on three indicators: handling time, degree of abnormal recovery, and recurrence alarm rate, and generate a handling performance statistical report.

10. The energy anomaly alarm pre-closed-loop control system based on hierarchical response according to claim 7, characterized in that, It also includes a visualization module, which provides three display terminals: a web management terminal, a mobile APP, and a large-screen monitoring terminal, and supports: The centralized display and cross-type correlation analysis of four types of alarms enable unified monitoring of the overall operating status of the energy system. Multi-dimensional statistical analysis reports showcasing alarm time distribution, type distribution, response time, and repeat alarm rate; Energy pipeline network GIS map visualization, intuitively showing the location and impact range of anomalies; Real-time tracking of disposal progress and full-process traceability and query of historical disposal records.

Images