A multi-source data driven power protection intelligent monitoring and early warning system

Abstract

The application is suitable for the technical field of power system, and provides a power protection intelligent monitoring and early warning system based on multi-source data driving, comprising: deploying data collection, processing, monitoring, alarming, forwarding and intelligent analysis modules through a safe partition architecture, and constructing a safe isolation transmission module comprising a first isolation channel, a second isolation channel and a reverse isolation channel; the data collection module accesses heterogeneous secondary equipment by automatically adapting multiple communication protocols; the data processing module intelligently classifies operation data into event type and state type, and determines transmission priority by classification; event type data enters a safe II area through the first isolation channel to realize real-time monitoring and alarming, and state type data enters a safe IV area through the second isolation channel to realize intelligent analysis; the intelligent analysis module outputs predictive early warning information to the real-time alarm module through the reverse isolation channel to optimize the alarm strategy; and the state perception, fault early warning and intelligent decision-making capability for power secondary equipment are improved.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a power protection intelligent monitoring and early warning system driven by multi-source data. Background Technology

[0002] In power systems, monitoring the operational status of secondary equipment such as relay protection and automatic safety devices, and performing fault analysis, are technical means to ensure the safe and stable operation of primary power equipment and the power grid. Currently, most monitoring methods rely on decentralized monitoring through local monitoring systems at substations or by uploading some key signals to a remote centralized control center via simple communication networks. Centralized monitoring of a single substation or a single type of equipment involves deploying data acquisition units to upload the equipment's action signals, alarm information, and waveform data to a backend system via a common network channel for display and alarm triggering.

[0003] In existing technologies, data acquisition is a single-mode access for specific devices or communication protocols, which cannot adapt to complex power production scenarios, including traditional hydropower stations and numerous new energy power plants. It is also difficult to unify the access and adapt to various secondary devices with different standards and protocols. Secondly, when data traffic surges, critical signals with extremely high real-time requirements, such as protection tripping and fault recording, may be overwhelmed by a large amount of routine status monitoring data, leading to delayed fault response. This prevents subsequent predictive analysis results from being effectively and securely fed back to and empowering the real-time monitoring and alarm decision-making process, causing the system as a whole to remain at the passive response level of post-event analysis, restricting the evolution of proactive early warning models, and resulting in a serious lack of intelligence. Summary of the Invention

[0004] This invention provides a power protection intelligent monitoring and early warning system based on multi-source data to solve the problems of insufficient evolution of active early warning modes and inadequate intelligence level.

[0005] This invention provides a power protection intelligent monitoring and early warning system based on multi-source data, including: a data acquisition module and a data processing module deployed in safety zone I, a real-time monitoring module, a real-time alarm module and a data forwarding module deployed in safety zone II, an intelligent analysis module deployed in safety zone IV, and a security isolation transmission module connecting the safety zones I, II and IV. The data acquisition module is used to collect operating data from secondary devices accessed from different communication networks through a variety of automatically adaptable communication protocols. The data processing module is used to classify the running data into event-type data and status-type data; and to classify the event-type data and status-type data according to preset classification rules to determine the priority of each type of data in cross-regional transmission. The secure isolation transmission module includes a first isolation channel, a second isolation channel, and a reverse isolation channel; wherein: The first isolation channel is deployed between Security Zone I and Security Zone II, and is used to transmit all operational data after the data processing module has been classified and graded to Security Zone II. The second isolation channel is deployed between Security Zone II and Security Zone IV to transmit status data from Security Zone II to Security Zone IV. The reverse isolation channel is deployed between security zone IV and security zone II, and is used to transmit the predictive early warning information output by the intelligent analysis module to the real-time alarm module in security zone II. The real-time monitoring module is used to receive the operation data transmitted through the first isolation channel and dynamically schedule display resources; The real-time alarm module is used to parse and pattern match the event-type data transmitted through the first isolation channel, generate an alarm signal when an anomaly matching a preset fault mode is detected, and receive predictive warning information transmitted through the reverse isolation channel to adjust the alarm strategy. The data forwarding module is used to process status data transmitted through the first isolation channel within Security Zone II, and forward it to the intelligent analysis module through the second isolation channel. The intelligent analysis module is used to perform intelligent analysis on the status data received and transmitted through the second isolation channel, output predictive early warning information, and transmit the predictive early warning information to the real-time alarm module of Security Zone II through the reverse isolation channel.

[0006] Furthermore, the data acquisition module is used to collect operational data from secondary devices accessed via different communication networks through automatically adaptable multiple communication protocols, including: Identify the communication protocol type used by the target secondary device, including standard communication protocols and non-standard proprietary protocols; When the protocol is identified as a standard communication protocol, the pre-built standard protocol parsing library is invoked to parse the received raw data packets and extract the runtime data. When a non-standard private protocol is identified, the corresponding protocol adaptation rules are matched according to the preset protocol feature library, and the received raw data packets are converted into runtime data in a unified data model format according to the matched protocol adaptation rules.

[0007] Furthermore, the data processing module is used to classify the operational data into event-type data and status-type data, including: Analyze the signal codes in the runtime data; Determine whether the signal code belongs to a predefined set of event-type signal codes, which includes the action output code, start command code, and fault trigger code of the protection element; If the signal code belongs to the event-type signal code set, then the running data is marked as event-type data; If the signal code does not belong to the event-type signal code set, then it is determined whether the signal code belongs to the predefined status-type signal code set, which includes analog measurement codes, switch status codes, and setpoint parameter codes. If the signal code belongs to the set of status signal codes, then the running data is marked as status data.

[0008] Furthermore, the data processing module classifies the event-type data according to preset classification rules, including: The corresponding protection action type is determined based on the signal code of the operational data marked as event-type data; The severity level of the event-type data is determined based on the type of protection action; wherein, a tripping action that causes equipment shutdown is the first severity level, protection activation is the second severity level, and protection alarm is the third severity level. A corresponding transmission queue instruction is generated based on the severity level; wherein, event-type data of the first severity level is assigned to the first transmission queue, and the data in the first transmission queue is preferentially transmitted to Security Zone II through the first isolation channel.

[0009] Furthermore, the data processing module classifies the state-type data according to preset classification rules, including: Identify the monitored device objects corresponding to the operational data marked as status data; According to the preset equipment criticality level table, the criticality level of the equipment object is obtained; wherein, the core equipment of the power plant is of the first criticality level, and the auxiliary equipment is of the second criticality level. If the device object corresponding to the status data is of the first critical level or the monitoring value in the status data exceeds the alarm threshold, it is allocated to the second transmission queue; wherein, the status data in the second transmission queue is preferentially forwarded to the security zone IV through the second isolation channel within the security zone II. Other status data is allocated to the third transmission queue and forwarded sequentially through the second isolation channel.

[0010] Furthermore, the first isolation channel is used to transmit all operational data from the data processing module after classification and grading to Security Zone II, including: Based on the running data output by the data processing module and the corresponding allocated transmission queue instructions, the running data is scheduled to Security Zone II; wherein, the transmission queue instructions include the instructions of the first transmission queue.

[0011] Furthermore, the second isolation channel is used to transmit status data from Security Zone II to Security Zone IV, including: The status data and the corresponding transmission queue instructions are received, and the status data is forwarded to security zone IV in the order indicated by the transmission queue instructions; wherein the transmission queue instructions include instructions for the second transmission queue and instructions for the third transmission queue.

[0012] Furthermore, the real-time alarm module is used to parse and pattern match the event-type data transmitted through the first isolation channel, and generate an alarm signal when an anomaly matching a preset fault mode is detected, including: A pre-set diagnostic process template is matched based on the event characteristics of the event-type data. The diagnostic process template corresponds to a preset fault mode and includes multiple diagnostic steps. The pre-set diagnostic process template includes at least two diagnostic steps, and different event characteristics of event-type data correspond to different combinations of diagnostic processes. The diagnostic steps in the diagnostic process template are executed sequentially to diagnose the event-type data. An alarm signal is generated when the diagnostic results meet the preset alarm conditions.

[0013] Furthermore, the real-time alarm module is used to parse and pattern match event-type data transmitted through the first isolation channel, and generates an alarm signal when an anomaly matching a preset fault mode is detected. It also includes: Receive predictive early warning information from the intelligent analysis module; Adjusting the diagnostic process template based on the predictive warning information includes adjusting the diagnostic steps and / or adjusting the judgment thresholds for the diagnostic steps; Use the revised diagnostic workflow template to diagnose event-based data.

[0014] Furthermore, the intelligent analysis module is used to intelligently analyze the status data received and transmitted through the second isolation channel, and output predictive early warning information, including: Temporal features are extracted from the received status data to construct a feature sequence reflecting the operating status of the equipment. The feature sequence is input into a pre-trained predictive analysis model to obtain the health assessment results and fault risk prediction results of the state-type data source equipment; Based on the health assessment results and fault risk prediction results, predictive early warning information is generated, which includes equipment identification, predicted fault type, risk level, and recommended monitoring strategies.

[0015] As can be seen from the above technical solutions, the present invention has the following advantages: This invention's system deploys data acquisition, processing, monitoring, alarming, forwarding, and intelligent analysis modules within a strict security partitioning architecture, and constructs a secure isolation transmission module including a first isolation channel, a second isolation channel, and a reverse isolation channel. The data acquisition module automatically adapts to multiple communication protocols to access heterogeneous secondary devices. The data processing module intelligently classifies operational data into event-based and status-based types, and performs hierarchical classification to determine transmission priorities. Event-based data enters Security Zone II via the first isolation channel for real-time monitoring and alarming, while status-based data enters Security Zone IV via the second isolation channel for intelligent analysis. Predictive early warning information output by the intelligent analysis module is fed back to the real-time alarm module in Security Zone II via the reverse isolation channel to dynamically optimize alarm strategies. This invention effectively solves the problems of difficult access to multi-source heterogeneous data and inaccurate data value identification through automatic protocol adaptation and unified classification and hierarchical classification, achieving comprehensive and standardized data acquisition. The multi-channel isolation transmission method based on business priorities ensures absolute priority and low latency for critical protection signals in cross-zone transmission, effectively improving the system's real-time response capability in fault conditions. This enhances the system's status perception, fault early warning, and intelligent decision-making capabilities for power secondary equipment. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the architecture of a power protection intelligent monitoring and early warning system based on multi-source data driving according to the present invention; Figure 2 This is a schematic diagram of the workflow of the data acquisition module and data processing module deployed in Security Zone I of this invention; Figure 3 This is a schematic diagram illustrating the workflow of the real-time monitoring module, real-time alarm module, and data forwarding module deployed in Security Zone II of this invention. Figure 4 This is a schematic diagram of the workflow of the intelligent analysis module deployed in Security Zone IV in this invention. Detailed Implementation

[0017] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “corresponding to,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] Example 1 Please see Figure 1The intelligent monitoring and early warning system for power protection based on multi-source data driven provided in this application includes: a data acquisition module and a data processing module deployed in Security Zone I, a real-time monitoring module, a real-time alarm module and a data forwarding module deployed in Security Zone II, an intelligent analysis module deployed in Security Zone IV, and a security isolation transmission module connecting Security Zone I, Security Zone II and Security Zone IV; the data acquisition module is used to acquire operating data from secondary equipment accessed from different communication networks through various automatically adapted communication protocols; the data processing module is used to classify the operating data into event-type data and status-type data; and classify the event-type data and status-type data according to preset classification rules to determine the priority of each type of data in cross-zone transmission; the security isolation transmission module includes a first isolation channel, a second isolation channel and a reverse isolation channel; wherein: the first isolation channel is deployed between Security Zone I and Security Zone II, and is used to transmit all operating data after classification and classification by the data processing module to Security Zone II; the second isolation channel is deployed in Security Zone II... Between Zone II and Zone IV, a reverse isolation channel is used to transmit status data from Zone II to Zone IV. A reverse isolation channel is deployed between Zone IV and Zone II to transmit predictive warning information output by the intelligent analysis module to the real-time alarm module in Zone II. A real-time monitoring module receives operational data transmitted via the first isolation channel and dynamically schedules display resources. A real-time alarm module parses and performs pattern matching on event data transmitted via the first isolation channel, generating an alarm signal when an anomaly matching a preset fault mode is detected. It also receives predictive warning information transmitted via the reverse isolation channel to adjust alarm strategies. A data forwarding module processes status data transmitted via the first isolation channel within Zone II and forwards it to the intelligent analysis module via the second isolation channel. The intelligent analysis module performs intelligent analysis on the received status data transmitted via the second isolation channel, outputs predictive warning information, and transmits the predictive warning information to the real-time alarm module in Zone II via the reverse isolation channel.

[0019] The principle of the system of the present invention will be described in detail below with reference to a specific scenario: The system of this invention has been deployed in the new energy centralized control center of a power generation group. This system is constructed in strict accordance with the safety protection regulations for power monitoring systems, and is divided into zones. In Safety Zone I, the data acquisition module connects to secondary equipment from different communication networks simultaneously via a front-end communication terminal deployed at the plant. The data processing module in Safety Zone I analyzes the acquired raw operational data, classifying it into event-based and status-based data according to the signal characteristics in the data frames. Simultaneously, this module prioritizes the two types of data according to the preset hierarchical rules of power protection business logic. For event-based data, it is classified according to the severity of protection actions; for status-based data, it is classified based on the importance of the equipment and the degree of numerical anomaly, generating corresponding transmission queue instructions. All data after hierarchical processing is transmitted from Safety Zone I to Safety Zone II through the first isolation channel in the safety isolation transmission module.

[0020] In Safety Zone II, the real-time monitoring module dynamically allocates display resources on the central control center's large screen based on priority, highlighting high-priority protection action events. Simultaneously, the real-time alarm module performs real-time parsing and pattern matching on incoming event-type data. When a sequence matching a preset typical fault mode is detected, it generates and issues audible and visual alarm signals to notify operators. On the other hand, the data forwarding module preprocesses status-type data through caching and packaging before sending it to Safety Zone IV via a second isolation channel.

[0021] In Safety Zone IV, the intelligent analysis module receives a continuous stream of massive amounts of status-based data. Its built-in time-series data analysis engine, based on a model trained on historical equipment operating data and a fault case library, performs trend analysis on the status data, identifies early signs of equipment degradation, and generates predictive warnings. These warnings are also securely and unidirectionally transmitted back to Safety Zone II via a dedicated reverse isolation channel deployed in the safety isolation transmission module. Upon receiving the predictive warnings from Zone IV, the real-time alarm module in Safety Zone II utilizes this information to optimize its real-time monitoring strategy. For example, when a warning indicates a potential risk in the excitation system of a unit, the alarm module automatically lowers the matching threshold for relevant abnormal signals of that unit and adds pre-diagnostic labels to potentially triggered alarms, enabling operators to proactively monitor, verify, and intervene, eliminating faults in their early stages.

[0022] Through the complete workflow described above, the system of this invention has successfully constructed a power protection intelligent monitoring and early warning system that spans safety zones I, II, and IV, integrates unified access to multi-source heterogeneous data, intelligent hierarchical scheduling of business, real-time monitoring, and in-depth analysis in the specific application of the power generation group's centralized control center. It has realized the intelligent evolution from post-analysis to real-time alarms and then to pre-warning.

[0023] Example 2 Please see Figure 2 The working principles of the data acquisition module and data processing module deployed in Security Zone I are described in detail below: In this embodiment, the data acquisition module is used to collect operating data from secondary devices accessed from different communication networks through multiple automatically adaptable communication protocols, including the following steps: 1. Identify the communication protocol type used by the target secondary device. Communication protocol types include standard communication protocols and non-standard proprietary protocols; 2. When the protocol is identified as a standard communication protocol, the pre-built standard protocol parsing library is invoked to parse the received raw data packets and extract the runtime data; 3. When a non-standard private protocol is identified, the corresponding protocol adaptation rules are matched according to the preset protocol feature library, and the received raw data packets are converted into runtime data in a unified data model format according to the matched protocol adaptation rules.

[0024] Specifically, the data acquisition module maintains a protocol feature library. This library stores the frame structure features of common standard protocols, as well as non-standard protocol feature templates obtained through reverse engineering of proprietary protocols from various equipment manufacturers or by summarizing them based on their technical manuals. When a raw data packet is received from a communication connection, this module extracts features such as its header and checksum method, and matches them with the protocol feature library. If a standard protocol is matched, the corresponding standard parser is loaded, and the packet is unpacked according to the protocol specification to extract structured data such as protection action signals and A-phase current measurements. If a non-standard protocol is matched, the corresponding adaptation script is loaded. This script defines the mapping rules from the raw binary packet to the system's unified data model, and after conversion, generates runtime data with a unified format and clear semantics.

[0025] In this embodiment, the data processing module is used to classify runtime data into event-type data and status-type data, including the following steps: 1. Analyze the signal codes in the runtime data; 2. Determine whether the signal code belongs to the predefined event-type signal code set. The event-type signal code set includes the protection element's action output code, start command code, and fault trigger code. 3. If the signal code belongs to the event-type signal code set, then the running data will be marked as event-type data; 4. If the signal code does not belong to the event-type signal code set, then determine whether the signal code belongs to the predefined status-type signal code set. The status-type signal code set includes analog measurement codes, digital status codes, and setpoint parameter codes. 5. If the signal code belongs to the set of status signal codes, then the running data will be marked as status data.

[0026] Specifically, in the communication protocol for secondary power equipment, each data point corresponds to a unique signal code. For example, YX_101 represents the switch position of line 1, and YC_205 represents the A-phase current of main transformer 2. The predefined set of event-type signal codes is constructed by summarizing all codes that may represent discrete changes, command execution, or abnormal triggering. This includes codes for all protection device trip outputs and all fault recorder start-up recordings. If the signal code of a running data belongs to this set, it indicates that the data represents an event that has already occurred, and therefore it is marked as event-type data. Conversely, the predefined set of status-type signal codes includes all codes representing continuously monitored quantities, such as analog measurement codes for current, voltage, and power, switch status codes for circuit breakers and disconnectors, and setting parameter codes for protection settings and control parameters. If a signal code belongs to this set, it indicates that the data represents the continuous state or fixed parameter of the equipment at a certain moment, and therefore it is marked as status-type data. This rigid classification rule based on signal code semantics is clear and executable, forming the basis for subsequent intelligent processing.

[0027] In this embodiment, the data processing module classifies event-type data according to preset classification rules, including the following sub-steps: 1. Determine the corresponding protection action type based on the signal code of the operation data marked as event-type data; 2. Determine the severity level of event-type data based on the type of protection action; among them, tripping actions that cause equipment shutdown are classified as the first severity level, protection activation as the second severity level, and protection alarms as the third severity level; 3. Generate corresponding transmission queue instructions based on the severity level; wherein, event-type data of the first severity level is assigned to the first transmission queue, and the data in the first transmission queue is preferentially transmitted to Security Zone II through the first isolation channel.

[0028] Specifically, the preset grading rules measure the immediate impact on power system operation. These rules first determine the corresponding protection action type based on the signal code. For example, the signal code BP_TRIP is mapped to bus protection tripping. Then, the rules define three severity levels: Level 1 severity refers to tripping actions that directly cause primary equipment to disconnect from the grid, having the greatest impact on system stability; Level 2 severity refers to activation signals where the protection device has started but has not yet issued a tripping signal, indicating that the equipment is within the fault's impact range or in an abnormal critical state; Level 3 severity refers to alarm signals that only reflect equipment abnormalities but will not immediately trigger a trip. Based on this, the system generates transmission queue instructions containing metadata including the target queue identifier. The first transmission queue is assigned the highest scheduling priority. Therefore, when event-type data is determined to be at Level 1 severity, the generated instructions will allocate it to the first transmission queue, ensuring that these most urgent signals are prioritized and quickly sent to Safety Zone II for processing during subsequent transmission through the first isolation channel, meeting the operational requirements for rapid fault response.

[0029] In this embodiment, the data processing module classifies the status data according to preset classification rules, including the following sub-steps: 1. Identify the monitored device objects corresponding to the operational data marked as status data; 2. Based on the pre-set equipment criticality level table, obtain the criticality level of the equipment; among them, the core equipment of the power plant is of the first criticality level, and the auxiliary equipment is of the second criticality level; 3. If the device object corresponding to the status data is of the first critical level or the monitoring value in the status data exceeds the alarm threshold, it shall be allocated to the second transmission queue; among which, the status data in the second transmission queue shall be preferentially forwarded to the security zone IV through the second isolation channel within the security zone II. 4. Other status data is allocated to the third transmission queue and forwarded sequentially through the second isolation channel.

[0030] Specifically, the equipment objects in status data are determined by both signal codes and equipment identifiers. For example, the signal code WINDING_TEMP and the equipment identifier TRANSFORMER_01 both point to the temperature of the No. 1 main transformer winding. The pre-defined equipment criticality level table is a mapping table predefined based on the equipment's function in the power plant's main electrical wiring and the scope of its impact during a power outage. For example, equipment such as generators, main transformers, and step-up station busbars, which could cause a plant-wide or large-scale power outage if they fail, are defined as first-level criticality equipment (core equipment), while equipment with a smaller impact, such as plant service transformers and standby transformers, are defined as second-level criticality equipment (auxiliary equipment). Alarm thresholds are safety limits pre-set based on equipment technical specifications, historical operating data, and industry standards. The classification rules stipulate that if any condition is met—either the equipment is a core device or the monitored value exceeds a limit—the status data has high analytical value or urgency, and therefore is allocated to the second transmission queue. This queue has priority in being processed and forwarded to safety zone IV within the data forwarding module of safety zone II. Regular status data that does not meet the above conditions is allocated to the third transmission queue. Data in the third transmission queue may be sent in batches during network idle periods or at scheduled times to optimize network bandwidth utilization. This hierarchical strategy ensures that analytical resources can focus on anomalies or potential risks in critical equipment.

[0031] Example 3 The working principle of the security isolation transmission module connecting Security Zone I, Security Zone II, and Security Zone IV is described in detail below: In this embodiment, the secure isolation transmission module is a set of hardware isolation devices built between different security zones to achieve strictly unidirectional and secure data transmission. Physically, this module typically consists of forward and reverse isolation devices deployed at the network boundaries of the security zones. These devices use non-network methods to achieve physical or protocol layer isolation and bridging, ensuring that even if the management information zone (Security Zone IV) suffers a network attack, it cannot penetrate to the production control zone (Security Zones II and I), while simultaneously meeting the compliance requirements for data exchange in the power monitoring system. The secure isolation transmission module includes three logical channels. The first isolation channel, deployed between Security Zone I and Security Zone II, is a unidirectional data transmission channel implemented by the forward isolation device, used to transmit operational data processed in Security Zone I to Security Zone II. The second isolation channel, deployed between Security Zone II and Security Zone IV, is another unidirectional data transmission channel implemented by a set of forward isolation devices, specifically used to transmit status data from Security Zone II to Security Zone IV. The reverse isolation channel, deployed between Security Zone IV and Security Zone II, is a one-way data feedback channel implemented by a reverse isolation device. It is specifically designed to transmit predictive early warning information generated in Security Zone IV to the real-time alarm module in Security Zone II. The workflow includes the following: 1. Based on the running data output by the data processing module and the corresponding allocated transmission queue instructions, schedule the running data to Security Zone II; wherein, the transmission queue instructions include the instructions of the first transmission queue.

[0032] 2. Receive status data and corresponding transmission queue instructions, and forward the status data to security zone IV in the order indicated by the transmission queue instructions; wherein, the transmission queue instructions include instructions for the second transmission queue and instructions for the third transmission queue.

[0033] Specifically, firstly, the transmitter in the Security Zone I channel receives the operating data and its corresponding transmission queue instruction from the data processing module. This transmission queue instruction contains metadata including data priority identifiers. The transmitter places the data into the corresponding transmission buffer queue according to the instruction. Data identified as being in the first transmission queue is placed in the high-priority buffer, while other data is placed in the normal-priority buffer. The scheduler within the channel operates strictly according to priority order, prioritizing the transmission of data in the high-priority buffer; only when the high-priority buffer is empty is data read from and sent from the normal-priority buffer. This priority-based preemptive scheduling method ensures that critical fault signals can traverse the isolation channel with minimal delay and reach the real-time monitoring and alarm module in Security Zone II.

[0034] Within Security Zone II, the data forwarding module maintains a logical transmission queue for status data to be sent to Security Zone IV. Upon receiving status data and its corresponding transmission queue instruction, the data forwarding module inserts the data into the designated position of the transmission queue according to the instruction. Specifically, data identified as belonging to the second transmission queue is inserted at the head of the queue—this represents high-priority status data from core devices or exceeding limits; while regular status data identified as belonging to the third transmission queue is inserted at the tail of the queue. The data forwarding module packages the status data according to this transmission queue from head to tail and sends it to the intelligent analysis module in Security Zone IV via the second isolation channel (forward isolation device). This ensures that highly urgent status data is analyzed first, optimizing the utilization efficiency of analysis resources.

[0035] The reverse isolation channel, acting as an independent, unidirectional feedback path, securely transmits predictive warning information generated by the intelligent analysis module in Security Zone IV back to the real-time alarm module in Security Zone II. This channel, also implemented using a dedicated reverse isolation device, ensures that the flow of information from the low-security zone to the high-security zone is controllable and cannot be reversed, thus enabling real-time monitoring services to benefit from intelligent analysis results while meeting the highest security standards.

[0036] Example 4 Please see Figure 3The working principles of the real-time monitoring module, real-time alarm module, and data forwarding module deployed in Security Zone II are described in detail below. In this embodiment, the real-time monitoring module receives and displays all operational data transmitted through the first isolation channel, serving as the core of the human-machine interface for operators to monitor the status of secondary equipment throughout the plant. Its function is to dynamically allocate display resources of the graphical user interface based on the priority information accompanying the operational data. For example, it automatically pops up and flashes the highest priority protection tripping event in the alarm window, while using different colors to distinguish and display different levels of status quantities, ensuring that operators can focus on the most critical system information immediately.

[0037] In this embodiment, the real-time alarm module is used to parse and pattern match the event-type data transmitted through the first isolation channel, and generate an alarm signal when an anomaly matching a preset fault mode is detected, including the following steps: 1. Match a pre-set diagnostic process template to the event characteristics of the event-type data. The diagnostic process template corresponds to a preset fault mode and includes multiple diagnostic steps. The pre-set diagnostic process template includes at least two diagnostic steps, and different event characteristics of the event-type data correspond to different combinations of diagnostic processes. 2. Perform each diagnostic step in the diagnostic workflow template sequentially to diagnose event-type data; 3. When the diagnostic results meet the preset alarm conditions, an alarm signal is generated.

[0038] Specifically, the preset fault modes are typical abnormal scenarios solidified based on power system fault analysis experience, such as line faults accompanied by correct protection operation, and switch failure leading to protection activation. For each fault mode, the system has a pre-set corresponding diagnostic process template, which is an ordered combination of a series of diagnostic steps. Common diagnostic steps include: a logic verification step to verify whether the protection action conforms to the preset logical coordination relationship; a timing verification step to check whether the timing of related switch changes and protection signals is reasonable; and a signal retrieval step to find the status and event-related signals of other related equipment in the same electrical bay at the same time. When event-type data arrives, the module matches the corresponding diagnostic process template according to its equipment type, signal type, and other characteristics, and executes the steps in the template sequentially. For example, it first checks whether there is a corresponding circuit breaker trip signal, then checks whether the trip timing is after the protection action, and finally retrieves whether the fault recording file has been generated. Each step generates a reliability score. The preset alarm condition is defined as the weighted sum of the scores of all key diagnostic steps exceeding a threshold, or the verification result of the key steps being a failure. When this condition is met, the module generates a structured alarm signal, which includes not only an anomaly description but may also include diagnostic conclusions and confidence levels.

[0039] In addition, this real-time alarm module also has the ability to dynamically optimize diagnostic strategies based on predictive information, specifically including: 1. Receive predictive early warning information from the intelligent analysis module; 2. Adjust the diagnostic process template based on predictive early warning information, including adjusting diagnostic steps and / or adjusting the judgment thresholds for diagnostic steps; 3. Use the revised diagnostic workflow template to diagnose event-based data.

[0040] Specifically, predictive early warning information is sent by the intelligent analysis module in Safety Zone IV through a reverse isolation channel, indicating a decline in the health of a certain device. Upon receiving this information, the alarm module adjusts the diagnostic process template related to the target device and its potential fault types. Adjustments include structural modifications, such as adding a specific vibration analysis step or oil chromatography data verification step to the diagnostic template for that device; and parametric adjustments, such as lowering the judgment thresholds of relevant diagnostic steps to make them more sensitive to abnormal signs. For example, if an early warning indicates a risk of inter-turn short circuit in main transformer No. 1, when the light gas alarm signal from main transformer No. 1 arrives, the alarm module will invoke an enhanced diagnostic template and lower the threshold value for requiring other corroborating signals for the light gas signal. This enables the system to respond more quickly and definitively to early, weak abnormal signals from high-risk equipment, realizing an evolution from passive alarms based on fixed rules to preventative alarms that combine real-time equipment health status.

[0041] Example 5 Please see Figure 4 The working principle of the intelligent analysis module deployed in Security Zone IV is described in detail below: In this embodiment, the intelligent analysis module is used to perform intelligent analysis on the status data received and transmitted through the second isolation channel, and output predictive early warning information, including the following steps: 1. Extract time-series features from the received status data to construct a feature sequence reflecting the operating status of the equipment; 2. Input the feature sequences into a pre-trained predictive analysis model to obtain the health assessment results and failure risk prediction results of the state-based data source equipment; 3. Based on the health assessment results and failure risk prediction results, generate predictive early warning information that includes equipment identification, predicted failure type, risk level, and recommended monitoring strategies.

[0042] Specifically, the intelligent analysis module is deployed on a high-performance server, utilizing a big data processing framework to store historical and real-time status data, and running various machine learning and statistical models for calculations. Status data is time-series data with timestamps; here, the raw data stream is transformed into a more representative feature sequence. First, statistics within a sliding time window are calculated to form a new feature sequence, including mean, standard deviation, kurtosis, and skewness. Vibration and current signals are transformed to extract time-domain features such as RMS value, peak factor, and impulse factor, as well as frequency-domain features such as the amplitude of major frequency components extracted through Fourier transform. Second, the slope of feature changes over a longer period is calculated using linear regression to form a trend feature sequence characterizing whether the equipment status is stable, slowly deteriorating, or accelerating. Finally, multiple monitoring quantities from the same equipment, after the above processing, are fused into a multi-dimensional, time-ordered feature vector sequence, which comprehensively reflects the equipment's operational behavior pattern over a period of time.

[0043] The pre-trained predictive analytics model is a mathematical model trained using machine learning methods based on historical data. Its training utilizes massive amounts of historical device state feature sequences as training samples, labeling the device with whether it will subsequently fail and what type of failure it will experience. Supervised learning algorithms are employed for training, enabling the model to learn to identify precursory patterns leading to specific failures from complex feature sequences. Multiple model types can be deployed in practical applications. The health assessment model is a regression model; it takes current and recent feature sequences as input and outputs a quantified health score, with lower scores indicating poorer overall device health. The failure risk prediction model is a sequence prediction model; it takes a feature sequence as input and outputs the probability of various specific failures occurring within a specified future time period. During online execution, the constructed current feature sequence is input into both the trained health assessment model and the failure risk prediction model, simultaneously obtaining real-time health assessment results and multiple failure risk prediction results for the device.

[0044] Finally, the model's numerical output is transformed into decision support information that operations and maintenance personnel can understand and execute. The system has a preset rule: when the health score falls below 60, or the predicted probability of any fault type exceeds 30%, an early warning is triggered. The generated predictive early warning information is a structured data object, containing at least the device identifier, predicted fault type, risk level, and suggested monitoring strategy. This early warning information is pushed to the real-time alarm module in Security Zone II via a reverse isolation channel.

[0045] It is understood that those skilled in the art can combine various implementation methods in the above embodiments under the guidance of the above examples to obtain technical solutions with multiple implementation methods.

[0046] 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.

Claims

1. A power protection intelligent monitoring and early warning system based on multi-source data, characterized in that, include: The system includes a data acquisition module and a data processing module deployed in Security Zone I, a real-time monitoring module, a real-time alarm module and a data forwarding module deployed in Security Zone II, an intelligent analysis module deployed in Security Zone IV, and a secure isolation transmission module connecting Security Zone I, Security Zone II and Security Zone IV. The data acquisition module is used to collect operating data from secondary devices accessed from different communication networks through a variety of automatically adaptable communication protocols. The data processing module is used to classify the running data into event-type data and status-type data; and to classify the event-type data and status-type data according to preset classification rules to determine the priority of each type of data in cross-regional transmission. The secure isolation transmission module includes a first isolation channel, a second isolation channel, and a reverse isolation channel; wherein: The first isolation channel is deployed between Security Zone I and Security Zone II, and is used to transmit all operational data after the data processing module has been classified and graded to Security Zone II. The second isolation channel is deployed between Security Zone II and Security Zone IV to transmit status data from Security Zone II to Security Zone IV. The reverse isolation channel is deployed between security zone IV and security zone II, and is used to transmit the predictive early warning information output by the intelligent analysis module to the real-time alarm module in security zone II. The real-time monitoring module is used to receive the operation data transmitted through the first isolation channel and dynamically schedule display resources; The real-time alarm module is used to parse and pattern match the event-type data transmitted through the first isolation channel, generate an alarm signal when an anomaly matching a preset fault mode is detected, and receive predictive warning information transmitted through the reverse isolation channel to adjust the alarm strategy. The data forwarding module is used to process status data transmitted through the first isolation channel within Security Zone II, and forward it to the intelligent analysis module through the second isolation channel. The intelligent analysis module is used to perform intelligent analysis on the status data received and transmitted through the second isolation channel, output predictive early warning information, and transmit the predictive early warning information to the real-time alarm module of Security Zone II through the reverse isolation channel.

2. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 1, characterized in that, The data acquisition module is used to collect operational data from secondary devices accessed through different communication networks using multiple automatically adaptable communication protocols, including: Identify the communication protocol type used by the target secondary device, including standard communication protocols and non-standard proprietary protocols; When the protocol is identified as a standard communication protocol, the pre-built standard protocol parsing library is invoked to parse the received raw data packets and extract the runtime data. When a non-standard private protocol is identified, the corresponding protocol adaptation rules are matched according to the preset protocol feature library, and the received raw data packets are converted into runtime data in a unified data model format according to the matched protocol adaptation rules.

3. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 1, characterized in that, The data processing module is used to classify the operational data into event-type data and status-type data, including: Analyze the signal codes in the runtime data; Determine whether the signal code belongs to a predefined set of event-type signal codes, which includes the action output code, start command code, and fault trigger code of the protection element; If the signal code belongs to the event-type signal code set, then the running data is marked as event-type data; If the signal code does not belong to the event-type signal code set, then it is determined whether the signal code belongs to the predefined status-type signal code set, which includes analog measurement codes, switch status codes, and setpoint parameter codes. If the signal code belongs to the set of status signal codes, then the running data is marked as status data.

4. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 3, characterized in that, The data processing module classifies the event-type data according to preset classification rules, including: The corresponding protection action type is determined based on the signal code of the operational data marked as event-type data; The severity level of the event-type data is determined based on the type of protection action; wherein, a tripping action that causes equipment shutdown is the first severity level, protection activation is the second severity level, and protection alarm is the third severity level. A corresponding transmission queue instruction is generated based on the severity level; wherein, event-type data of the first severity level is assigned to the first transmission queue, and the data in the first transmission queue is preferentially transmitted to Security Zone II through the first isolation channel.

5. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 3, characterized in that, The data processing module classifies the status data according to preset classification rules, including: Identify the monitored device objects corresponding to the operational data marked as status data; According to the preset equipment criticality level table, the criticality level of the equipment object is obtained; wherein, the core equipment of the power plant is of the first criticality level, and the auxiliary equipment is of the second criticality level. If the device object corresponding to the status data is of the first critical level or the monitoring value in the status data exceeds the alarm threshold, it is allocated to the second transmission queue; wherein, the status data in the second transmission queue is preferentially forwarded to the security zone IV through the second isolation channel within the security zone II. Other status data is allocated to the third transmission queue and forwarded sequentially through the second isolation channel.

6. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 4, characterized in that, The first isolation channel is used to transmit all operational data from the data processing module after classification and grading to Security Zone II, including: Based on the running data output by the data processing module and the corresponding allocated transmission queue instructions, the running data is scheduled to Security Zone II; wherein, the transmission queue instructions include the instructions of the first transmission queue.

7. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 5, characterized in that, The second isolation channel is used to transmit status data from Security Zone II to Security Zone IV, including: The status data and the corresponding transmission queue instructions are received, and the status data is forwarded to security zone IV in the order indicated by the transmission queue instructions; wherein the transmission queue instructions include instructions for the second transmission queue and instructions for the third transmission queue.

8. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 1, characterized in that, The real-time alarm module is used to parse and pattern match event-type data transmitted through the first isolation channel, and generate an alarm signal when an anomaly matching a preset fault mode is detected, including: A pre-set diagnostic process template is matched based on the event characteristics of the event-type data. The diagnostic process template corresponds to a preset fault mode and includes multiple diagnostic steps. The pre-set diagnostic process template includes at least two diagnostic steps, and different event characteristics of event-type data correspond to different combinations of diagnostic processes. The diagnostic steps in the diagnostic process template are executed sequentially to diagnose the event-type data. An alarm signal is generated when the diagnostic results meet the preset alarm conditions.

9. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 8, characterized in that, The real-time alarm module is used to parse and pattern match event-type data transmitted through the first isolation channel, and generates an alarm signal when an anomaly matching a preset fault mode is detected. It also includes: Receive predictive early warning information from the intelligent analysis module; Adjusting the diagnostic process template based on the predictive warning information includes adjusting the diagnostic steps and / or adjusting the judgment thresholds for the diagnostic steps; Use the revised diagnostic workflow template to diagnose event-based data.

10. The intelligent monitoring and early warning system for power protection based on multi-source data as described in claim 1, characterized in that, The intelligent analysis module is used to perform intelligent analysis on the status data received and transmitted through the second isolation channel, and output predictive early warning information, including: Temporal features are extracted from the received status data to construct a feature sequence reflecting the operating status of the equipment. The feature sequence is input into a pre-trained predictive analysis model to obtain the health assessment results and fault risk prediction results of the state-type data source equipment; Based on the health assessment results and fault risk prediction results, predictive early warning information is generated, which includes equipment identification, predicted fault type, risk level, and recommended monitoring strategies.

Images