A centralized station intelligent power distribution operation and maintenance integrated management system

Abstract

The application discloses a kind of collection station wisdom power distribution operation integrated management and control systems, including field perception layer, edge computing layer and cloud platform layer.Field perception layer collects the multi-source operating parameters of equipment, power supply and environment in station.Edge computing layer carries out protocol analysis, preprocessing and local analysis on multi-source heterogeneous data, to generate standardized data.In the cloud platform layer, the data governance module governs the standardized data, forming multi-theme asset data;The digital twin module constructs a digital twin that synchronously maps with the physical collection station;The intelligent analysis module performs fault root cause analysis and operation strategy simulation verification based on the twin;The application service module provides monitoring, alarm and work order management, and other operation and maintenance functions.The system realizes deep perception, intelligent diagnosis, strategy simulation and closed-loop control of the operation status of the collection station through cloud-edge-end collaboration and digital twin technology, effectively improving operation and maintenance efficiency, decision accuracy and system security.

Description

Technical Field

[0001] This invention relates to the field of power system automation and intelligent operation and maintenance technology, and in particular to an integrated management and control system for intelligent distribution operation and maintenance of a collection station. Background Technology

[0002] With the continuous increase in the proportion of new energy power generation, the stability of the operation and the efficiency of operation and maintenance management of collection stations, which are key nodes for the centralized access of distributed power sources and the aggregation of electricity, are crucial to the safety of the distribution network. Currently, the operation and maintenance management of these stations generally relies on the deployment of various independent monitoring systems. For example, a Supervisory Control and Data Acquisition (SCADA) system deployed within the station is responsible for collecting electrical quantity information, an independent video security system is used for image monitoring, and equipment management and work order processes may rely on another production management system. These systems, to a certain extent, achieve real-time monitoring and basic control of the main equipment and environment within the station, constituting the main technical support for operation and maintenance work.

[0003] However, the existing technology model based on discrete systems has gradually revealed several limitations in practical applications. First, the subsystems are typically built by different vendors at different times, using varying communication protocols and data formats. This leads to data source isolation within the site, creating "information silos" that cannot provide a unified and consistent data foundation for advanced analytics applications. Second, maintenance work heavily relies on human experience. Judgments of equipment anomalies or faults are mostly based on threshold alarms and post-event analysis, lacking the ability to deeply trace the root causes of faults. Furthermore, it is impossible to conduct pre-simulation verification of the safety and effectiveness of operational adjustments or fault handling strategies before implementing actual control measures. This results in delays and uncertainties in the maintenance decision-making process, making it difficult to shift from passive response to proactive prevention.

[0004] Therefore, facing modern data collection stations that are increasingly large in scale, with more complex equipment, and closely integrated with the power grid, traditional operation and maintenance technologies are no longer sufficient to meet the demands for efficient, precise, and intelligent management and control. The industry urgently needs an integrated intelligent management and control system that can break down data barriers, deeply integrate multi-source information, and possess in-depth analysis and simulation decision-making capabilities. Summary of the Invention

[0005] To improve existing methods and systems, a smart power distribution operation and maintenance integrated management and control system for collection stations is provided, including: The field sensing layer is configured to collect and aggregate multi-source operating parameters of electrical equipment, distributed power sources, and the environment within the station. The edge computing layer, which communicates with the field perception layer, is configured to receive and process multi-source operating parameters. The processing includes parsing, preprocessing, and local intelligent analysis of data from different manufacturers and communication protocols in the multi-source operating parameters to generate standardized data. The cloud platform layer, which communicates with the edge computing layer via network, is configured to receive and process standardized data. The cloud platform layer includes: The data governance module is configured to integrate, govern, and store the received standardized data to form standardized multi-theme asset data; The digital twin module is configured to build and drive a digital twin that is synchronously mapped with the physical collection station based on multi-theme asset data; The intelligent analysis module interacts with the digital twin module and is configured to perform root cause analysis of faults and simulation verification of operating strategies based on the digital twin. The application service module is configured to call the services of the digital twin module and the intelligent analysis module to provide operation and maintenance management functions, including monitoring, alarms and work order management.

[0006] In some embodiments, the digital twin integrates an electrical topology model for characterizing the actual connection relationships of various electrical devices within the station.

[0007] In some embodiments, the intelligent analysis module includes a fault tracing analysis unit and an operation strategy simulation unit; wherein... The fault tracing and analysis unit is configured to locate faults and analyze their causes based on digital twins; The operational strategy simulation unit is configured to simulate and verify control strategies or contingency plans in a digital twin.

[0008] In some embodiments, the fault tracing analysis unit is further configured to: when an alarm signal is received, invoke the electrical topology model, starting from the alarm device, traverse along the topology connection relationship, determine the associated monitoring points on the upstream power supply side and downstream load side of the alarm device; extract the operating data of the alarm device and the set of associated monitoring points before and after the alarm period for comparison and analysis, so as to locate the root fault source and generate an analysis report.

[0009] In some embodiments, the operation strategy simulation unit is configured to: load candidate operation control strategies or fault handling plans into a digital twin; simulate and calculate the changes in voltage, current, and power state variables of key electrical nodes after the execution of the operation control strategy or fault handling plan based on the electrical topology model and equipment parameter model; compare the simulated voltage, current, and power state variables with preset safety and stability constraints, and output compliance verification results and optimized control command sequences.

[0010] In some embodiments, the intelligent analysis module further includes an energy efficiency optimization unit, which is configured to: calculate the power generation efficiency of each distributed power unit and the network loss under different operating paths based on historical and real-time data; generate a draft scheduling strategy including equipment switching and power allocation with the overall system energy efficiency as the objective function; and send the draft scheduling strategy to the operation strategy simulation unit for safety and feasibility verification.

[0011] In some embodiments, the edge computing layer's data processing for photovoltaic power generation units includes: converting the per-unit power data uploaded by the photovoltaic inverter, based on the rated capacity, into the actual power value; calculating the sum of the current data of each photovoltaic string belonging to the same combiner box as the combiner box output current; and calculating the deviation of each branch current from the average value to generate a current imbalance index for judging branch abnormalities.

[0012] In some embodiments, the data governance module includes a protocol adaptation unit, which is configured to: call a pre-configured Modbus-TCP protocol adapter and an IEC 61400-25 protocol adapter to parse the uplink data packets of the photovoltaic inverter and the wind turbine, respectively; map the parsed data points according to a predefined unified metadata model, assign them unique identifiers and register them in the metadata repository, and then store them in the corresponding subject database.

[0013] In some embodiments, the cloud platform layer further includes a case library module, configured to store cases consisting of historical fault characteristics and handling records; the fault tracing and analysis unit is further configured to: when generating a fault analysis report, match similar historical cases from the case library module based on the current fault characteristics, and associate the handling records of the matched historical cases as reference suggestions with the analysis report for output.

[0014] In some embodiments, the field sensing layer includes wireless temperature sensing units deployed at switch cabinet contacts and cable joints; the digital twin module is further configured to dynamically bind and fuse the temperature data monitored in real time by the wireless temperature sensing units with the spatial position of the corresponding three-dimensional model of the device in the digital twin, and trigger an overheat warning when the temperature exceeds a preset threshold.

[0015] This application provides an integrated intelligent power distribution operation and maintenance management system for collection stations, which adopts a collaborative architecture of a field perception layer, an edge computing layer, and a cloud platform layer. The field perception layer is responsible for collecting multi-source operating parameters, the edge computing layer performs protocol parsing and local preprocessing of heterogeneous data, and the cloud platform layer includes data governance, digital twin, intelligent analysis, and application service modules. This solution uses the data governance module to form standardized multi-theme asset data, which drives a digital twin synchronized with the physical station. The intelligent analysis module uses this twin to achieve fault root cause localization and operational strategy simulation verification. The application service module ultimately provides integrated monitoring, alarm, and work order functions. The beneficial effects are that, through cloud-edge-device collaboration and digital twin technology, it solves the problem of multi-source data fusion, realizes the transformation of operation and maintenance decision-making from passive response to proactive prevention, and improves the accuracy of fault analysis, the security of strategy verification, and the overall efficiency of operation and maintenance management. Attached Figure Description

[0016] Figure 1 This is a system architecture diagram of the intelligent power distribution operation and maintenance integrated management and control system for collection stations proposed in this invention.

[0017] In the diagram: 1. On-site perception layer; 2. Edge computing layer; 3. Cloud platform layer; 31. Data governance module; 32. Digital twin module; 33. Intelligent analysis module; 34. Application service module. Detailed Implementation

[0018] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0019] Firstly, this application provides an integrated management and control system for intelligent power distribution operation and maintenance of collection stations, such as... Figure 1 As shown, it includes: The field perception layer 1 is configured to collect and aggregate multi-source operating parameters of electrical equipment, distributed power sources and environment within the station; Edge computing layer 2, which is connected to field perception layer 1, is configured to receive and process multi-source operating parameters. The processing includes parsing, preprocessing and local intelligent analysis of data from different manufacturers and communication protocols in the multi-source operating parameters to generate standardized data. Cloud platform layer 3, which is connected to edge computing layer 2 via network communication, is configured to receive and process standardized data. Cloud platform layer 3 includes: The data governance module 31 is configured to integrate, govern, and store the received standardized data to form standardized multi-theme asset data; Digital twin module 32 is configured to build and drive a digital twin that is synchronously mapped with the physical collection station based on multi-theme asset data; The intelligent analysis module 33 interacts with the digital twin module 32 and is configured to perform root cause analysis of faults and simulation verification of operating strategies based on the digital twin. Application service module 34 is configured to call the services of digital twin module 32 and intelligent analysis module 33 to provide operation and maintenance management functions including monitoring, alarm and work order management.

[0020] This embodiment provides a specific implementation method for an integrated intelligent power distribution operation and maintenance management system for a data collection station. The system deploys a wide-ranging sensor network on-site, forming the on-site sensing layer 1. These sensors are directly installed on equipment such as transformers, switchgear, photovoltaic inverters, and wind turbines, as well as in the station environment, to continuously collect multi-source operating parameters such as voltage, current, power, temperature, equipment status signals, and meteorological data. This raw data is the foundation for all system analyses, but its diverse sources and formats necessitate effective integration.

[0021] To address the issue of data heterogeneity, the system deploys an edge computing layer 2 locally within the site. This layer, comprised of industrial-grade gateways or edge servers, is primarily responsible for receiving raw data streams from devices from different vendors. It integrates multiple industrial communication protocol stacks (such as Modbus-TCP, IEC 104, and IEC 61400-25), enabling automatic identification and parsing of messages from different protocols. Beyond protocol parsing, edge computing layer 2 performs preprocessing tasks such as data cleaning, timestamp alignment, and unit conversion, and can run lightweight analysis algorithms, such as calculating the imbalance of photovoltaic string currents and performing preliminary limit checks. After these processes, the originally chaotic raw data is transformed into a standardized data stream with a uniform format and reliable quality before being uploaded to the cloud. This significantly reduces the processing burden on the cloud and improves data timeliness.

[0022] The cloud platform layer 3 is the system's "intelligent brain." The data governance module 31 first integrates and governs the massive amounts of standardized data received. It categorizes, correlates, and reorganizes the data according to themes such as equipment, power generation, topology, and environment, and stores it in a time-series database or data lake, ultimately forming standardized, easily accessible multi-theme asset data. These high-quality data assets are the foundation for building advanced applications. The digital twin module 32 then uses this asset data to construct a dynamically synchronized digital twin in virtual space that corresponds one-to-one with the physical collection station. This twin not only includes a three-dimensional appearance model of the equipment but, more importantly, integrates precise electrical wiring topology and real equipment parameters, enabling the virtual world to accurately map the real-time status and connectivity of the physical world.

[0023] Based on this vivid and accurate digital twin, the intelligent analysis module 33 can unleash its advanced capabilities. When a fault occurs, the analysis module can invoke the topology model within the twin, starting from the alarm point and automatically tracing the operating data of upstream and downstream related equipment along the electrical path to quickly locate the root cause and achieve precise root cause analysis. When adjustments to operating modes or the development of emergency repair plans are needed, the analysis module can simulate and verify the strategies in this "safety sandbox" of the digital twin, pre-imagine the system state changes after the strategy execution, and automatically verify whether it meets safety procedures, thereby outputting verified and safe optimization instructions. Finally, the application service module 34 provides maintenance personnel with a unified operating interface. By invoking the visualization services of the twin and the decision services of the intelligent analysis module 33, it transforms the complex analysis results into intuitive monitoring screens, clear alarm information, structured analysis reports, and executable maintenance work orders, thus completing a complete control loop from perception, analysis, decision-making to execution.

[0024] In summary, this embodiment creatively integrates digital twin technology into the entire operation and maintenance process of a data collection station through a three-layer architecture of on-site perception, edge computing, and cloud collaboration. This system effectively solves the fundamental problems of data dispersion and system isolation in traditional models, achieving deep fusion and unified management of multi-source data. Its core value lies in transforming the operation and maintenance mode from passive response relying on experience to proactive early warning and precise decision-making based on data and model simulation verification, leveraging a digital twin synchronized with the physical entity. This significantly improves fault handling efficiency, operational strategy security, and the overall intelligence level of operation and maintenance management, laying a solid technical foundation for achieving unmanned operation and predictive maintenance of data collection stations.

[0025] In some embodiments, the digital twin integrates an electrical topology model for characterizing the actual connection relationships of various electrical devices within the station.

[0026] In its implementation, to ensure that the digital twin transcends mere 3D appearance and truly possesses a calculable and analyzable "soul," a crucial step in this embodiment is the deep integration of an electrical topology model to accurately represent the actual connections of various electrical devices within the station. This model is not simply a graphical representation; rather, based on the station's design drawings, equipment ledgers, and on-site survey data, it is constructed using graphical modeling tools or specialized software. This model represents the transformers, circuit breakers, disconnectors, busbars, lines, photovoltaic combiner boxes, inverters, and other equipment within the station, according to their actual electrical connections (such as series, parallel, star, or delta connections), forming a networked data model containing nodes and branches. This model clarifies the logical path of electrical energy flow, for example, indicating the complete electrical link from the photovoltaic array through the combiner box and inverter, to the step-up transformer and grid connection point, enabling the virtual twin to accurately reflect the energy and information transmission relationships within the physical entity.

[0027] To achieve accurate construction and dynamic maintenance of this model, various technical methods can be employed in practice. One common approach is to utilize existing digital design files, such as BIM (Building Information Modeling) or SCD (System Configuration Description) files, for parsing and conversion. This automatically extracts equipment and connection information from the primary wiring diagram and transforms it into topology data that can be recognized and calculated by a computer. Another method involves scanning electronic nameplates or QR codes on equipment using mobile terminals after on-site construction or renovation, allowing for manual confirmation and recording of connections between devices, thereby calibrating the theoretical model on-site. A concrete example is that when troubleshooting a feeder tripping fault, the system can immediately invoke the electrical topology model, automatically analyze all loads and distributed power supply nodes connected downstream of the feeder, and quickly obtain real-time current and protection action signals from these related devices. This provides maintenance personnel with crucial clues to define the scope of the fault's impact, entirely dependent on the precise connection knowledge provided by the topology model.

[0028] The integration of the electrical topology model lays the foundation for advanced analysis functions and directly brings significant benefits. In fault analysis scenarios, the intelligent analysis module 33 can use this model to perform "topology tracing". When a circuit breaker trips due to an alarm, the system no longer views the event in isolation, but automatically locates all upstream and downstream devices of the circuit breaker based on the model, and simultaneously retrieves the waveform data and event sequence records of these related points before and after the fault for comparison and analysis. This efficiently distinguishes whether the tripping is caused by a fault in the downstream line or by a malfunction of the circuit breaker itself, achieving a leap from "alarm point monitoring" to "precise location of system-level fault root causes". In operational strategy simulation verification, this model enables power flow calculations and protection coordination checks within the digital twin. For example, before executing the instruction to "switch some photovoltaic power generation from one lightly loaded line to another," this operation can be simulated in the twin based on the topology model. The changes in load rates of each line and transformer after the power flow transfer can be calculated, and it can be automatically checked whether it will cause any equipment overload or protection malfunction. This realizes the transformation of operational decision-making from "experience-based judgment" to "simulation pre-play and safety verification." Therefore, the digital twin integrating the electrical topology model essentially constructs a precise logical mapping connecting the physical world and the information world, and is a core enabling element driving operation and maintenance management towards intelligence, precision, and prevention.

[0029] In some embodiments, the intelligent analysis module 33 includes a fault source analysis unit and an operation strategy simulation unit; wherein... The fault tracing and analysis unit is configured to locate faults and analyze their causes based on digital twins; The operational strategy simulation unit is configured to simulate and verify control strategies or contingency plans in a digital twin.

[0030] In the specific deployment and application of the system, the core functions of the intelligent analysis module 33 are achieved collaboratively through its two internal specialized units—the fault source analysis unit and the operation strategy simulation unit—which together endow the digital twin with advanced capabilities for dynamic analysis and decision support. The fault source analysis unit was designed to change the inefficient traditional operation and maintenance model that relies on manual point-by-point troubleshooting. Its working principle is that when the monitoring system receives an alarm signal, such as an excessively high temperature rise in a box-type transformer, this unit immediately activates and invokes the digital twin. It does not examine the transformer model in isolation, but first automatically identifies the upstream power supply lines (which may include specific photovoltaic power generation units or feeders) supplying the transformer and all the loads downstream of them, based on the electrical topology model integrated in the twin. Subsequently, the unit synchronously extracts and correlates detailed operating data of these upstream and downstream devices before and after the alarm time, such as fluctuations in upstream power output, sudden changes in downstream load, and the transformer's own historical load curves and cooling system status. Through this cross-device and cross-time sequence correlation comparison, the unit can intelligently infer the root cause, such as whether it is due to the start-up of a large downstream device causing long-term overload, or due to its own cooling fan failure, and automatically generate a structured analysis report that clearly presents the fault chain to the operation and maintenance personnel.

[0031] Meanwhile, the operation strategy simulation unit focuses on virtual verification before actual actions are taken, acting as a "pre-operation sandbox" for operation and maintenance decisions. This unit has a wide range of applications. For example, before planning a power outage maintenance of a section of busbar within the station, maintenance personnel need to formulate a load transfer strategy on a digital twin. The operation strategy simulation unit then loads and executes this strategy in the twin based on a real-time snapshot of the current system status. It simulates and calculates the redistribution of power flow across the entire station after the strategy is executed, based on the electrical topology model and the precise parameter models of the equipment, predicting voltage changes at key nodes, load rates of each line, and whether the settings of protection devices will be triggered. If the simulation results show that the transferred load will cause an overload on another line, the unit will immediately issue a warning and may attempt to provide adjustment suggestions, such as phased operation or switching to backup power, until a safe and feasible optimal solution is simulated. This is equivalent to conducting a comprehensive, zero-risk stress test before actual operation.

[0032] This architecture, which concretizes intelligent analysis into two highly specialized units, delivers clear and significant benefits. The fault tracing and analysis unit, through its topology-based, data-driven correlation analysis method, elevates fault diagnosis from "point" observation to "system" insight, greatly reducing the time from alarm to root cause location and minimizing secondary problems caused by misjudgments, thus improving the accuracy and efficiency of operations and maintenance. The operation strategy simulation unit, by constructing a high-fidelity virtual verification environment, transforms key decisions such as adjusting operating modes and formulating maintenance plans from relying on personal experience and static procedures to scientific decisions based on dynamic simulation and quantitative evaluation. This fundamentally reduces the risk of misoperation, ensures the safety of system operation, and enables more economical operation strategies through iterative simulation optimization. Together, these two units transform the entire operations and maintenance system from a passive, reactive model of post-event remediation to an intelligent closed-loop management model capable of proactively predicting risks, simulating and optimizing solutions, and accurately eliminating faults. This is the core technological support for achieving unmanned and intelligent operations and maintenance at the collection station.

[0033] In some embodiments, the fault tracing analysis unit is further configured to: when an alarm signal is received, invoke the electrical topology model, starting from the alarm device, traverse along the topology connection relationship, determine the associated monitoring points on the upstream power supply side and downstream load side of the alarm device; extract the operating data of the alarm device and the set of associated monitoring points before and after the alarm period for comparison and analysis, so as to locate the root fault source and generate an analysis report.

[0034] In actual operation and maintenance scenarios of data collection stations, the causes of faults are often cascading and interconnected. The fault tracing and analysis unit in this embodiment is designed to precisely address this challenge. Its core mechanism is that when the system captures an alarm signal from any device, the unit does not examine the alarm point in isolation, but immediately places it within the networked context of the entire electrical system. It first invokes a pre-defined, precise electrical topology model in the digital twin to identify the alarm device as the origin of the logical analysis. Subsequently, the unit automatically performs a bidirectional traversal search along the electrical connections defined in the topology model: one path traces upstream to the power supply side, exploring the real-time status of the buses, lines, and even distributed power sources supplying it; the other path extends downstream to the load side, sorting out all load devices within its power supply range. This process automatically determines a dynamic "set of associated monitoring points," which includes all key monitoring locations on the electrical path that have direct or indirect energy interaction with the alarm point.

[0035] To achieve this intelligent traceability, the system relies on a digital twin to map and bind nodes in the topology model to measurement points in the real-time database during technology implementation. For example, when a branch of a photovoltaic combiner box reports a "severe current imbalance" alarm, the fault tracing analysis unit will initiate the above process. Starting from that branch, it immediately correlates with other normal branches of the combiner box, the total output of the combiner box, and the DC and AC side parameters of the connected inverter through the topology model. Simultaneously, it may also correlate with environmental monitoring data from the same photovoltaic subarray. Then, the unit will extract the alarm branch, the total output of the combiner box, the inverter power, and the irradiance data for the same period in parallel, performing comparative analysis within a time window before and after the alarm occurred. By analyzing the time-series relationship between the differences in current in each branch and the changes in total output and irradiance, it is possible to efficiently determine whether the imbalance is caused by a single branch module failure, the difference in overall irradiance changes between branches, or an abnormality in the inverter acquisition unit. This allows for precise identification of the root cause of the fault, which may be the failure of a certain photovoltaic string rather than a problem with the entire combiner box or inverter. The system also automatically generates an analysis report that identifies the root cause and includes a chain of data evidence.

[0036] This intelligent tracing mechanism, based on topological traversal and correlation data comparison, has brought revolutionary benefits. It completely changes the traditional troubleshooting model that relies on manual experience and rote memorization, transforming fault location from a time-consuming and error-prone "skill" into a fast and accurate "automated analysis process." Its most direct benefit is a significant reduction in average fault diagnosis and recovery time. A deeper value lies in its ability, through systematic correlation analysis, to effectively identify the true "root fault" and the resulting "secondary alarms," ​​avoiding piecemeal and ineffective solutions and improving the accuracy of operational decisions. Ultimately, the standardized and structured analysis reports generated by this unit not only guide current efficient handling but also serve as valuable knowledge reserves, continuously feeding back into and optimizing the system's diagnostic rules and thresholds. This enables the entire operational system to continuously learn and evolve, thereby significantly improving the reliability of the power supply to the collection station.

[0037] In some embodiments, the operation strategy simulation unit is configured to: load candidate operation control strategies or fault handling plans into a digital twin; simulate and calculate the changes in voltage, current, and power state variables of key electrical nodes after the execution of the operation control strategy or fault handling plan based on the electrical topology model and equipment parameter model; compare the simulated voltage, current, and power state variables with preset safety and stability constraints, and output compliance verification results and optimized control command sequences.

[0038] The core value of the operation strategy simulation unit lies in constructing a high-fidelity, zero-risk "digital testbed" for various operational decisions at the collection station. In actual operation, when planned switching operations, adjustments to generator output, or activation of emergency response plans for certain faults need to be performed, traditional methods heavily rely on dispatching procedures and human experience, posing a risk of misoperation due to insufficient consideration. The workflow of this unit begins with accurately loading candidate strategies or plans described in digital format into the digital twin. For example, a switching operation ticket detailing "close switch 101 and simultaneously disconnect switch 102," or an automated handling logic that "automatically activates backup power and isolates fault areas in sections after detecting a line protection action."

[0039] The key to its technical implementation lies in simulation calculation. The unit does not perform simple logical deduction, but rather drives a digital twin to simulate the dynamic process of the system after the strategy is executed, based on its inherent electrical topology model and precise equipment parameter models (such as transformer impedance, line resistance and reactance, circuit breaker operating time, etc.). It uses power system simulation algorithms to calculate the time-varying curves of state variables such as voltage, current, and power at various key electrical nodes in the network after the operation occurs. For example, when simulating a strategy to transfer part of the photovoltaic load from one line to another, the unit can calculate the potential power flow impact at the moment of transfer, the long-term load rate of the target line after commissioning, and the fluctuations in the relevant bus voltage. After completing the simulation, the unit automatically compares the obtained series of state variables with a preset library of safety and stability constraints, including equipment rated capacity, protection settings, voltage acceptable range, system stability limits, etc. If the simulation results show that all key parameters meet the constraints, the output will be a "compliant" verification result and a sequence of control commands that can be directly issued. If any risk of exceeding the limit is found (such as a transformer being overloaded), a detailed warning report will be output, and the original strategy can be adjusted and iterated in combination with the optimization algorithm until a safe and feasible optimization solution is generated.

[0040] Through the aforementioned mechanisms, the operational strategy simulation unit delivers multiple profound benefits. The most direct effect is a significant improvement in operational safety. By pre-exposing potential risks in a virtual space, it effectively prevents real power grid accidents caused by human error or policy flaws, transforming "post-event remediation" into "pre-event prevention." Secondly, it enhances the economy and efficiency of operation and maintenance. Simulation allows for the verification and optimization of operational sequences, reducing unnecessary power outages and equipment operation frequency, and exploring more economical operating methods. From a scientific decision-making perspective, this unit transforms the traditional decision-making model based on clauses and experience into a precise decision-making model based on quantitative data simulation and automatic rule verification, ensuring that every important operation is based on evidence and verifiable data. In the long run, it enables the formulation and application of complex system cascading operations and adaptive recovery strategies, significantly enhancing the resilience of collection stations in coping with complex operating conditions and sudden failures, representing a key indicator of a leap forward in the level of intelligent operation and maintenance.

[0041] In some embodiments, the intelligent analysis module 33 further includes an energy efficiency optimization unit, which is configured to: calculate the power generation efficiency of each distributed power unit and the network loss under different operating paths based on historical and real-time data; generate a draft scheduling strategy including equipment switching and power allocation with the overall system energy efficiency as the objective function; and send the draft scheduling strategy to the operation strategy simulation unit for safety and feasibility verification.

[0042] In the daily operation of a power collection station, improving power generation revenue and reducing internal losses are important economic objectives. However, traditional methods often rely on fixed rules or isolated adjustments, making it difficult to achieve optimal overall energy efficiency at the system level. Therefore, this embodiment introduces an energy efficiency optimization unit specifically within the intelligent analysis module 33. Its function is to proactively seek the optimal solution for the overall station's energy efficiency based on data-driven principles. This unit continuously receives historical and real-time data streams provided by the data governance module 31. It not only calculates the instantaneous power generation efficiency of individual photovoltaic inverters or wind turbines but also analyzes their performance degradation trends through long-term data. Simultaneously, based on the electrical topology model and line parameters in the digital twin, it dynamically calculates the network loss differences under different operating modes (e.g., power transmission from photovoltaic array A to the main transformer via line 1 versus via line 2), thereby accurately mapping the cost of energy flow within the station.

[0043] Based on this dynamic cost map, the core task of the energy efficiency optimization unit is to perform optimization calculations. It uses "optimal overall system energy efficiency" as its explicit objective function, which comprehensively considers maximizing power generation and minimizing network losses. The unit runs optimization algorithms to generate specific draft scheduling strategies. For example, during the midday hours when sunlight is intense and photovoltaic output is at full capacity, the draft might recommend putting all inverters in operation, but optimizing the switching of some power originally transmitted via long-distance, high-impedance lines to shorter paths, and appropriately lowering the power point tracking target of inverters connected to inefficient strings in the shade to prevent their inefficient operation from dragging down overall efficiency. Alternatively, during the evening when load is low, the draft might recommend preemptively disconnecting some transformers located on the edge of the grid with low transmission power but relatively high no-load losses. All these specific suggestions involving equipment switching and power allocation are presented in the strategy draft in the form of a clear sequence of instructions.

[0044] However, purely energy efficiency optimization may introduce new safety risks. Therefore, this unit is designed with a crucial closed loop: it automatically sends the generated draft scheduling strategy to the operational strategy simulation unit. The operational strategy simulation unit immediately performs rigorous safety and feasibility checks on this draft in a digital twin, simulating whether the system power flow exceeds limits, whether the voltage is stable, and whether the protection devices malfunction after executing these optimization instructions. Only strategies that pass simulation verification are approved for implementation. This closed-loop mechanism of "optimization proposal - simulation verification" is the ingenious design feature of this embodiment.

[0045] The introduction of an energy efficiency optimization unit and the establishment of the aforementioned closed loop have brought about several significant benefits. First, it elevates energy efficiency management from static, isolated equipment monitoring to dynamic, system-level collaborative optimization, continuously tapping into the potential for energy conservation and consumption reduction, and directly improving the economic efficiency of collection station operation. Second, by forcibly importing the optimization draft into the safety simulation process, it fundamentally eliminates the potential risk of "sacrificing safety for economic gain," achieving a balance between the two major goals of economy and safety. Finally, this enables the system to automatically and intelligently adapt to weather changes, load fluctuations, and equipment status changes, achieving adaptive optimization operation at the energy efficiency level, significantly reducing reliance on manual experience and judgment, and promoting the in-depth development of operation and maintenance management towards refinement and intelligence.

[0046] In some embodiments, the data processing of the photovoltaic power generation unit by the edge computing layer 2 includes: converting the per-unit power data uploaded by the photovoltaic inverter based on the rated capacity into the actual power value; calculating the sum of the current data of each photovoltaic string belonging to the same combiner box as the output current of the combiner box; and calculating the deviation of the current of each branch relative to the average value to generate a current imbalance index for judging branch abnormalities.

[0047] In the operation and monitoring of the collection station, photovoltaic power generation units present unique challenges to data processing due to their large number of devices, dense data points, and high real-time requirements. This embodiment clarifies the key preprocessing process of edge computing layer 2 for this type of data. Its primary task is to perform "de-normalization" processing on the data uploaded by the inverters. For the sake of equipment compatibility and communication efficiency, photovoltaic inverters often upload their output power in per-unit value based on their rated capacity. For example, an inverter with a rated power of 100kW uploads a value of 0.85. After receiving this data, the edge computing node will query the inverter's nameplate parameters from the local or cloud configuration library, perform a simple per-unit inverse transformation operation (0.85 * 100kW), and restore it to the actual power value of 85kW. This step is crucial because it ensures that the power data from hundreds or thousands of inverters of different brands and capacities have a unified and comparable physical dimension in the subsequent analysis and statistics of the system, laying a reliable data foundation for accurate assessment of the overall power generation capacity and calculation of equivalent utilization hours, among other advanced applications.

[0048] While standardizing the power values, Edge Computing Layer 2 simultaneously performs another important diagnostic calculation—combiner box branch current imbalance analysis. A photovoltaic combiner box typically collects current from a dozen to several dozen photovoltaic strings. Edge computing nodes collect the current data of all branches belonging to the same combiner box in real time. First, they calculate the sum of these current values ​​as the total output current of the combiner box, a key operational status variable. More importantly, it further calculates the average value of all branch currents, and then calculates the absolute or relative percentage deviation of each branch current from this average value. For example, for a 16-branch combiner box, if the calculation finds that 15 of the currents are around 8A, while the 16th current is only 1A, the system will immediately generate an extremely high current imbalance index. This index is no longer the raw data, but a derived characteristic quantity that directly characterizes the health status of the equipment.

[0049] This targeted processing at the data source yields direct and significant benefits. Unifying the per-unit power output to actual values ​​enables the system to seamlessly and accurately perform total power generation statistics, power prediction, and cross-device performance benchmarking, eliminating analytical errors caused by inconsistent data formats and improving the consistency and availability of global data. Real-time generation of current imbalance indicators frees maintenance personnel from the massive amounts of raw data, allowing them to focus directly on abnormal signals. This enables the system to automatically and instantly detect performance degradation in power generation units caused by string faults, shading, loose connections, or PID effects, achieving online diagnosis and early warning of the photovoltaic array's health status. This changes the traditionally lagging model of relying on periodic manual inspections or post-event power generation comparisons to discover problems, allowing maintenance personnel to accurately locate abnormal branches and intervene promptly, thereby effectively reducing power generation losses and improving the operational efficiency and profitability of photovoltaic assets. These two edge-side data processing methods together constitute the efficient and accurate "front-end sensing nerves" of the intelligent operation and maintenance system.

[0050] In some embodiments, the data governance module 31 includes a protocol adaptation unit, which is configured to: call a pre-configured Modbus-TCP protocol adapter and an IEC 61400-25 protocol adapter to parse the uplink data packets of the photovoltaic inverter and the wind turbine respectively; map the parsed data points according to a predefined unified metadata model, assign them unique identifiers and register them in the metadata repository, and then store them in the corresponding topic database.

[0051] In the actual system integration of the data collection station, how to efficiently and reliably access the massive amounts of operational data generated by photovoltaic inverters and wind turbine generators from numerous manufacturers and of varying models is the first technical hurdle that must be overcome in building an integrated management and control platform. In this embodiment, the data governance module 31 addresses this challenge specifically through its built-in protocol adaptation unit. This unit pre-configures dedicated parsing adapters for mainstream industry communication protocols at the software level, such as a Modbus-TCP protocol adapter for parsing most photovoltaic inverter data, and an IEC 61400-25 protocol adapter for parsing wind turbine data conforming to international standards. When raw data packets arrive from physical devices, the protocol adaptation unit automatically identifies the protocol type and calls the corresponding adapter for decoding. Taking Modbus-TCP as an example, the adapter accurately extracts the original values ​​of each data point from the binary message based on the pre-configured device point table (such as register address 40001 representing phase A current); for the IEC 61400-25 protocol, the adapter parses its complex, data model-based structured message to obtain rich information such as wind turbine speed, blade angle, and generator speed.

[0052] However, simply parsing the raw values ​​is far from sufficient; the key lies in achieving data standardization and semantic uniformity. To this end, the protocol adaptation unit performs a crucial second step: data mapping and registration. All parsed data points, whether it's the inverter's "DC voltage" or the wind turbine's "instantaneous power," are transformed according to a unified metadata model predefined at the system design stage. This model specifies the data naming, units, precision, and data types shared across the entire system. For example, the "power" field reported by inverters from different manufacturers may be uniformly mapped to "active_power," with the unit forcibly converted to "kilowatts" (kW). Each standardized data point is assigned a unique identifier within the entire system, carrying descriptive information such as its associated equipment, physical location, and signal type, and is registered as a complete "data asset" in the central metadata repository. After registration, the data is then categorized and stored in the corresponding time-series database or relational database according to its business theme (such as "power generation" or "equipment status").

[0053] This series of highly automated data processing procedures has brought fundamental and crucial benefits. It fundamentally solves the long-standing "hard nut to crack" problem of data access in smart power plant construction. Through software-based and configurable protocol adaptation, it replaces the traditional point-to-point integration model, which requires customized development, is costly, and difficult to maintain, greatly improving the system's access speed and flexibility to heterogeneous devices. More importantly, by mandating unified metadata mapping and centralized registration, it transforms the chaotic and ambiguous "raw data" into standardized data assets with clear formats and semantics, completely breaking down the "data semantic barriers" between different devices and systems. This provides high-quality data raw materials for the accurate construction of the upper-layer digital twin, ensuring the consistency and reliability of model parameters in the twin with the real world. It also enables the subsequent intelligent analysis module 33 to conduct cross-device and cross-system correlation analysis, performance benchmarking, and advanced computing, forming a solid foundation for the entire integrated management and control system's data value chain, transforming "data-driven operation and maintenance" from a concept into a stable and reliable engineering practice.

[0054] In some embodiments, the cloud platform layer 3 further includes a case library module, configured to store cases consisting of historical fault characteristics and handling records; the fault tracing and analysis unit is further configured to: when generating a fault analysis report, match similar historical cases from the case library module based on the current fault characteristics, and associate the handling records of the matched historical cases as reference suggestions with the analysis report for output.

[0055] During the continuous operation and accumulation of intelligent operation and maintenance systems, a wealth of valuable fault handling experience is generated. However, this experience is often scattered across different personnel or fragmented records, making it difficult to reuse systematically. The case library module added in this embodiment aims to transform this tacit knowledge into explicit assets that can be directly accessed and learned by the system. This module serves as a dedicated storage and retrieval center in the cloud platform layer 3. It is not merely a simple database, but rather a standardized case library that stores historical fault characteristics and effective handling records according to a standardized structure. Each case is refined and includes several key dimensions: the main characteristics of the fault occurrence (e.g., "transformer high temperature alarm, accompanied by a sudden increase in load, ambient temperature normal"), snapshots of relevant equipment operating data before and after the fault, verified root cause analysis conclusions, and the final successful handling steps and results. These cases accumulate over time, forming an ever-expanding knowledge graph surrounding the operation and maintenance of this collection station.

[0056] When the fault tracing and analysis unit is processing a newly occurring fault and generating an analysis report, it automatically activates collaboration with the case library module. The unit extracts the salient features of the current fault, such as alarm type, affected equipment type, and associated environmental parameters, and transforms these features into a standardized query vector. Subsequently, the system uses this query vector to perform intelligent matching and retrieval in the case library. The matching process may employ rule-based, similarity calculation, or even simple machine learning models to identify several historically most similar cases. For example, when the system encounters a fault of "frequent start-stop of photovoltaic inverters," the fault tracing and analysis unit, while performing its own topology and data analysis based on real-time data, will match similar historical cases from the case library. The matching results may show that two similar faults in the past were caused by an abnormality in the DC-side voltage sampling module and by a momentary drop in AC-side grid voltage. The detailed features, analysis process, and handling records of these historical cases will be automatically linked to the real-time fault analysis report being generated, providing valuable reference suggestions to maintenance personnel.

[0057] The introduction of a case study library module and its deep integration with intelligent analysis bring cognitive enhancement and continuous evolution capabilities that traditional systems lack. Its most direct benefit is a significant improvement in the quality and speed of operational decision-making. It's equivalent to equipping every operations and maintenance (O&M) personnel with a tireless and experienced "digital expert" who can immediately provide historical case handling strategies and successful experiences when new faults occur, assisting in rapid decision-making and shortening fault assessment time, especially beneficial for less experienced personnel. A deeper value lies in its ability to accumulate and transfer O&M knowledge from individual experience to organizational assets, preventing knowledge loss due to personnel turnover. This allows the entire team's O&M capabilities to be built upon historical accumulation, achieving iterative progress "standing on the shoulders of giants." In the long run, the continuously enriched case study library provides a high-quality training data foundation for future advanced fault self-diagnosis and handling suggestions, enabling the entire O&M system to possess the potential for self-learning and evolution, thereby driving a fundamental shift in the O&M model from "human experience-driven" to "data and knowledge-driven."

[0058] In some embodiments, the field sensing layer 1 includes wireless temperature sensing units deployed at switch cabinet contacts and cable joints; the digital twin module 32 is further configured to dynamically bind and fuse the temperature data monitored in real time by the wireless temperature sensing units with the spatial position of the corresponding three-dimensional model of the device in the digital twin, and trigger an overheat warning when the temperature exceeds a preset threshold.

[0059] In the daily operation and maintenance of the collection station, overheating of electrical connection points is one of the main hidden dangers causing fires and equipment damage. However, traditional manual handheld infrared temperature measurement or periodic inspection methods have significant shortcomings such as large monitoring blind spots and poor real-time performance. To address this, this embodiment has made key enhancements to the field sensing layer 1 by deploying temperature sensing units based on wireless communication technology at critical locations prone to heat generation, such as circuit breaker contacts, disconnector switch connections, busbar connection points, and outdoor cable terminations. These sensors typically use battery power and low-power wide-area network technologies such as LoRa and ZigBee, enabling long-term, real-time monitoring of the precise temperature of the contacts without affecting equipment operation or altering the original structure. The data is then wirelessly transmitted to the aggregation node within the station, greatly expanding the coverage and continuity of temperature monitoring.

[0060] After acquiring these real-time temperature data streams, the digital twin module 32 undertakes the crucial task of making them "visualized" and "intelligent." The module doesn't simply treat the temperature data as a list of independent values; instead, it performs a vital spatial mapping operation: it dynamically binds the logical number or physical address of each wireless temperature sensing unit to its specific spatial location on the corresponding high-precision 3D device model within the digital twin. For example, it associates the sensor ID installed on the "contact of switch 101 in the No. 1 incoming line cabinet" with the precise contact location on the 3D model of that switch within the twin. Based on this, the module achieves fused data display. It can intuitively render the real-time temperature values ​​on the accurate location of the corresponding device model on the 3D twin screen, using heat maps, color patches, or dynamic digital labels. Maintenance personnel can not only see the overall appearance of the switch cabinet on the monitoring screen but also directly see which specific contact point inside has a temperature of 42℃ (displayed in green) and which point has reached 85℃ (displayed in striking red). Meanwhile, the module sets a customizable warning threshold for each monitoring point. When the temperature data at any point continues to exceed the threshold (such as 90°C for ordinary connection points or 75°C for cable joints), the system will automatically trigger a multi-level overheat warning, highlighting and flashing the alarm point in the 3D screen, and generating a warning event containing the specific location and temperature curve, which will be pushed to the alarm center of the application service module 34.

[0061] The implementation of this technical solution has brought multi-layered and tangible benefits. First, it overcomes the industry pain point of difficulty in real-time monitoring of temperatures inside enclosed cabinets and at dense contact points, achieving "all-weather, no-blind-spot" perception of hidden thermal hazards and significantly advancing the prevention of accidents. Second, by precisely binding and integrating temperature data with the spatial location of a 3D model, it transforms abstract digital alarms into highly intuitive and precisely positioned visual warnings. Maintenance personnel can "at a glance" pinpoint the exact equipment and location of overheating, saving time spent comparing complex drawings and physical objects, greatly improving emergency response speed. Finally, the systematic overheating early warning mechanism transforms the maintenance mode from "periodic spot checks" to "continuous monitoring and automatic early warning," enabling timely detection of early temperature rises caused by increased contact resistance, overload, or internal faults. This provides a clear basis for planned maintenance and proactive intervention, effectively preventing serious accidents such as equipment burnout or fires caused by overheating, and significantly improving the safe operation level and risk management capabilities of the collection station equipment.

[0062] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0063] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0064] 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, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart power distribution operation and maintenance integrated management and control system for a collection station, characterized in that, include: The field sensing layer (1) is configured to collect multi-source operating parameters of electrical equipment, distributed power supply and environment in the collection station; The edge computing layer (2) is communicatively connected to the field perception layer (1) and is configured to receive and process the multi-source operating parameters. The processing includes parsing, preprocessing and local intelligent analysis of data from different manufacturers and communication protocols in the multi-source operating parameters to generate standardized data. A cloud platform layer (3), which is network-communication connected to the edge computing layer (2), is configured to receive and process the standardized data. The cloud platform layer (3) includes: The data governance module (31) is configured to integrate, govern and store the received standardized data to form standardized multi-theme asset data; The digital twin module (32) is configured to construct and drive a digital twin that is synchronously mapped with the physical collection station based on the multi-theme asset data; The intelligent analysis module (33) interacts with the digital twin module (32) and is configured to perform root cause analysis of faults and simulation verification of operating strategies based on the digital twin. The application service module (34) is configured to call the services of the digital twin module (32) and the intelligent analysis module (33) to provide operation and maintenance management functions including monitoring, alarm and work order management.

2. The system according to claim 1, characterized in that, The digital twin integrates an electrical topology model that characterizes the actual connection relationships of various electrical devices within the station.

3. The system according to claim 2, characterized in that, The intelligent analysis module (33) includes a fault tracing analysis unit and an operation strategy simulation unit; wherein, The fault tracing and analysis unit is configured to locate faults and analyze their causes based on the digital twin; The operation strategy simulation unit is configured to perform simulation verification of control strategies or contingency plans in the digital twin.

4. The system according to claim 3, characterized in that, The fault tracing and analysis unit is also configured to: when an alarm signal is received, invoke the electrical topology model, take the alarm device as the starting point, traverse along the topology connection relationship, determine the associated monitoring points on the upstream power supply side and downstream load side of the alarm device; extract the operating data of the alarm device and the set of associated monitoring points before and after the alarm period for comparison and analysis, so as to locate the root fault source and generate an analysis report.

5. The system according to claim 4, characterized in that, The operation strategy simulation unit is configured to: load candidate operation control strategies or fault handling plans into the digital twin; based on the electrical topology model and equipment parameter model, simulate and calculate the changes in voltage, current, and power state variables of key electrical nodes after the execution of the operation control strategy or the fault handling plan; compare the simulated voltage, current, and power state variables with preset safety and stability constraints, and output compliance verification results and optimized control command sequences.

6. The system according to claim 5, characterized in that, The intelligent analysis module (33) also includes an energy efficiency optimization unit, which is configured to: calculate the power generation efficiency of each distributed power unit and the network loss under different operating paths based on historical and real-time data. With the overall system energy efficiency as the objective function, a draft scheduling strategy including equipment switching and power allocation is generated, and the draft scheduling strategy is sent to the operation strategy simulation unit for safety and feasibility verification.

7. The system according to claim 1, characterized in that, The edge computing layer (2) performs the following data processing on the photovoltaic power generation unit: it back-calculates the per-unit power data uploaded by the photovoltaic inverter based on the rated capacity into the actual power value; it calculates the sum of the current data of each photovoltaic string belonging to the same combiner box as the output current of the combiner box, and calculates the deviation of each branch current from the average value to generate a current imbalance index for judging branch abnormalities.

8. The system according to claim 1, characterized in that, The data governance module (31) includes a protocol adaptation unit, which is configured to: call the preset Modbus-TCP protocol adapter and IEC 61400-25 protocol adapter to parse the uplink data packets of the photovoltaic inverter and the wind turbine respectively; map the parsed data points according to the predefined unified metadata model, assign a unique identifier and register them to the metadata repository, and then store them in the corresponding topic database.

9. The system according to claim 3, characterized in that, The cloud platform layer (3) also includes a case library module, which is configured to store cases consisting of historical fault characteristics and handling records; the fault tracing analysis unit is also configured to: when generating a fault analysis report, match similar historical cases from the case library module based on the current fault characteristics, and associate the handling records of the matched historical cases as reference suggestions with the analysis report for output.

10. The system according to claim 1, characterized in that, The field perception layer (1) includes wireless temperature sensing units deployed at switch cabinet contacts and cable joints; the digital twin module (32) is also configured to dynamically bind and fuse the temperature data monitored in real time by the wireless temperature sensing unit with the spatial position of the corresponding three-dimensional model of the equipment in the digital twin, and trigger an overheat warning when the temperature exceeds a preset threshold.

Images