A BIM-based power electromechanical equipment safety monitoring method, system and device

Abstract

The application relates to a BIM-based power electromechanical equipment safety monitoring method, system and device, and belongs to the technical field of construction supervision. The method fuses multiple models such as three-dimensional geometry and thermal physical properties, constructs a digital twin body mapping a physical power system, and synchronizes multiple-source heterogeneous data such as equipment shell temperature, load current, environmental parameters and protection signals. The inverse problem of heat conduction is solved by a fixed cycle combined with a thermal physical model to inversely calculate the temperature field and heat source power in the equipment, the thermal risk is evaluated by comparing the insulation heat-resistant threshold, and a predictive maintenance work order is generated when the threshold is met; early fault models are matched by real-time data processing to realize fault early warning, protection signals and switch states are listened to, and the self-healing process is triggered when the conditions are met. The electrical topology model is relied on to determine the fault interval, an electrical isolation scheme is generated and optimized, control instructions are issued after safety verification, fault isolation and non-fault section power supply recovery are completed, the operation and maintenance labor cost is effectively reduced, and the fault loss is reduced.

Description

Technical Field

[0001] This application relates to the technical field of construction supervision, and in particular to a BIM-based method, system and equipment for safety monitoring of electrical and mechanical equipment. Background Technology

[0002] With the continuous advancement of new power system construction, the proportion of new energy sources such as wind power and photovoltaics is constantly increasing, and the number of distributed power sources is growing exponentially. The operating environment of power electromechanical equipment is becoming increasingly complex, placing higher demands on the accuracy, real-time performance, and intelligence level of equipment safety monitoring. As a core component of the power system, the operating status of power electromechanical equipment directly affects the stability, security, and reliability of the power grid. Once a fault occurs, it will not only cause huge economic losses but may also trigger public safety incidents such as large-scale power outages.

[0003] Currently, safety monitoring of power electromechanical equipment mainly relies on a combination of single-sensor data collection and manual inspection. Specifically, single-type monitoring devices such as temperature and current sensors are deployed at key parts of the equipment to collect basic operating parameters like equipment casing temperature and load current. In some scenarios, environmental sensors are additionally deployed to collect environmental parameters such as temperature and humidity. Simultaneously, regular manual inspections are conducted to check the equipment's appearance, switch status, and protection signals to supplement the limitations of sensor monitoring. However, this approach lacks the ability to accurately identify early-stage faults, making early warning impossible. When a fault occurs, maintenance personnel must conduct on-site investigations to determine the fault area, which is heavily influenced by human experience and can easily lead to delayed fault handling. This not only increases maintenance labor costs but may also exacerbate potential damage from the fault. Summary of the Invention

[0004] To reduce maintenance manpower costs and minimize failure losses, this application provides a BIM-based method, system, and equipment for safety monitoring of power electromechanical equipment.

[0005] Firstly, this application provides a BIM-based safety monitoring method for power electromechanical equipment, employing the following technical solution: A BIM-based safety monitoring method for power electromechanical equipment includes: By integrating the three-dimensional geometric model, thermophysical property model, real-time electrical topology model, and control logic model of power equipment, a digital twin that maps the physical power system is constructed. Acquire multi-source heterogeneous data of the physical power system and synchronize the multi-source heterogeneous data to the digital twin; the multi-source heterogeneous data includes the collected equipment casing temperature, load current, environmental parameters, and the acquired protection signals and switch status; With a preset first fixed period, in the digital twin, based on the synchronized device casing temperature, load current and environmental parameters, and combined with the thermophysical property model, the inverse problem of heat conduction is solved to deduce the predicted temperature field and heat source power inside the device. The predicted temperature field is compared with the preset equipment insulation heat resistance threshold to assess the equipment thermal risk level. When the warning or alarm level is reached, a predictive maintenance work order containing equipment location and predicted hot spot information is generated and output. The multi-source heterogeneous data synchronized to the digital twin is processed in real time and matched with a preset early fault model. If a match is found, a corresponding fault warning is generated and output. At the same time, the protection signals and switch status are monitored in real time. When an event that meets the preset electrical fault judgment conditions is detected, the self-healing process is triggered. After the autonomous healing process is triggered, the fault range is determined based on the real-time electrical topology model in the digital twin, and multiple feasible electrical isolation schemes are generated. The optimal scheme is determined from the multiple electrical isolation schemes based on preset evaluation criteria. After safety logic verification, a sequence of control commands is generated and sent to the physical system to perform fault isolation and restore power supply to non-faulty sections.

[0006] By adopting the above technical solutions and constructing a digital twin through the integration of multi-dimensional models, a deep mapping and dynamic synchronization between the physical power system and the digital space is achieved. Through real-time access and analysis of multi-source heterogeneous data, it is possible not only to accurately invert the internal temperature field and heat source power of equipment by solving the inverse problem of heat conduction, thus identifying thermal risks in advance and generating predictive maintenance work orders to eliminate potential equipment failures at an early stage, but also to achieve real-time fault early warning and autonomous response by matching early fault models with real-time data and monitoring protection signals and switch states. After a fault occurs, the real-time electrical topology model of the digital twin quickly determines the fault area and generates the optimal isolation scheme. After safety verification, control commands are automatically issued to complete fault isolation and power restoration, improving the efficiency and safety of power system fault handling. This reduces the labor costs and fault losses of power equipment operation and maintenance while improving the stability of power system operation.

[0007] Optionally, the steps of solving the inverse problem of heat conduction by combining the aforementioned thermophysical property model to deduce the predicted temperature field and heat source power inside the equipment specifically include: The average temperature of the housing is calculated based on the readings of multiple temperature sensors arranged on the housing. Estimate the baseline heating power of the equipment based on the collected effective value of the load current; A heat conduction equation is established with the internal hot spot temperature of the equipment as the unknown core and the average temperature of the outer shell and the ambient temperature as boundary conditions. An optimization algorithm is used to iteratively adjust the assumed value of the internal heat source power to minimize the error between the shell temperature distribution calculated based on the assumed value and the measured temperature distribution using the heat conduction equation. The corresponding heat source power and the calculated internal temperature at this point are the inversion results.

[0008] Optionally, the steps for assessing the thermal risk level of equipment and generating predictive maintenance work orders specifically include: The temperature rise margin is obtained by calculating the difference between the internal hot spot temperature and the maximum allowable temperature of the equipment insulation material. Calculate the rate of change of the internal hotspot temperature between the current cycle and the previous cycle to obtain the temperature rise rate; When the temperature rise margin is lower than the first threshold or the temperature rise rate is positive and exceeds the set rate, it is determined to be a warning level; When the temperature rise margin is lower than a lower second threshold or the internal hot spot temperature reaches the allowable limit, it is determined to be an alarm level; In the three-dimensional geometric model of the digital twin, the corresponding equipment model is highlighted with color according to the risk level, and a structured work order containing the equipment's unique identifier, predicted hot spot locations, suggested maintenance items, and recommended time windows is automatically generated.

[0009] By adopting the above technical solution and determining the temperature rise margin and rate of temperature rise from two dimensions, a refined and dynamic perception of the equipment's thermal state is achieved. This not only captures changes in the insulation performance margin of the equipment but also promptly detects abnormal temperature rise trends, making thermal risk warnings and alarms more proactive and accurate. Furthermore, by using a digital twin's three-dimensional geometric model for color-coded risk level rendering, abstract thermal risk data is transformed into intuitive visual information, reducing the difficulty of identification for maintenance personnel.

[0010] Optionally, the step of determining the optimal solution from multiple electrical isolation solutions based on preset evaluation criteria specifically includes: Centered on the fault point, all feasible electrical isolation schemes are simulated in parallel within the digital twin. For each electrical isolation scheme, online simulation is performed and the comprehensive loss function value is calculated. The electrical isolation scheme with the smallest comprehensive loss function value is selected as the optimal scheme.

[0011] Optionally, the step of performing online simulation and calculating the comprehensive loss function value for each of the electrical isolation schemes specifically includes: The switching operation of the corresponding electrical isolation scheme is simulated and executed in the digital twin copy to update the electrical topology connection state; Fast power flow estimation is performed based on the updated electrical topology to obtain the reconstructed current values ​​of each branch and the voltage values ​​of each node. Based on the reconstructed current value and combined with the equipment thermal model, the new heat generation power and hot spot temperature of the key equipment can be quickly estimated, and potential thermal risk indicators can be calculated. The load loss caused by the corresponding electrical isolation scheme is statistically analyzed, and the load loss index is calculated by combining the preset load importance weights. The number of switches required to be operated and the physical distance of the operation path for the corresponding electrical isolation scheme are statistically analyzed, and the operation complexity index is calculated. Based on the load loss index, operational complexity index, and potential thermal risk index, a weighted summation algorithm is applied to calculate the comprehensive loss function value of the corresponding electrical isolation scheme.

[0012] By adopting the above technical solution, and simulating switch operations and updating the topology in a digital twin copy, the operational risks that may arise from testing in the physical system are avoided. Based on the updated topology, rapid power flow estimation is performed, and combined with equipment thermal models to extrapolate the heating and hotspot temperatures of key equipment, enabling proactive prediction of potential thermal risks of the isolation scheme. This allows for early avoidance of equipment overload or overheating issues caused by topology reconfiguration. By statistically analyzing power loss loads and combining them with importance weights to calculate load loss indicators, both power supply reliability and the priority differences of different loads are taken into account. Furthermore, the statistics on the number of operated switches and path distances quantify the implementation cost of the scheme, making the evaluation more comprehensive. This not only improves the accuracy and comprehensiveness of electrical isolation scheme evaluation but also minimizes operating costs and equipment operational risks while ensuring power supply reliability.

[0013] Optionally, the security monitoring method further includes: After the equipment completes maintenance according to the predictive maintenance work order, the actual location and severity of the defects in the maintenance record are obtained. The actual defect information is compared with the hotspot locations and temperatures predicted by the digital twin inversion, and the prediction deviation is calculated. Based on the prediction deviation, the corresponding parameters of the thermophysical property model in the digital twin are adjusted in reverse to narrow the gap between the virtual model and the physical entity.

[0014] Optionally, the security monitoring method further includes: Record the actual handling results of each failure event, including the actual isolation range, load recovery time, and whether it caused subsequent problems; The actual handling results are compared and analyzed with the comprehensive loss function values ​​of each candidate electrical isolation scheme during online simulation in the digital twin; When the actual optimal solution differs from the solution selected in the simulation, or when there is a systematic deviation between the simulation evaluation index and the actual handling result, the weight coefficients of load loss, operational complexity, and potential thermal risk in the comprehensive loss function are iteratively adjusted.

[0015] By adopting the above technical solution and using an iterative optimization mechanism based on feedback from actual handling results, a complete closed loop is constructed, from virtual simulation to physical execution and then to model correction. When there is a deviation between the actual optimal solution and the solution selected in the simulation, or when there are systematic differences between the simulation evaluation indicators and the actual results, the weight coefficients of each indicator in the comprehensive loss function are iteratively adjusted. This allows the evaluation logic of the simulation model to continuously align with the actual operating scenario, gradually eliminating the error between virtual and reality, thereby improving the accuracy and reliability of subsequent fault handling solution selection.

[0016] Optionally, the security monitoring method further includes: After generating the predictive maintenance work order or the fault warning, the thermal risk level or fault warning level information of the corresponding equipment is synchronized to the decision logic of the autonomous healing process. When calculating the load loss index of the comprehensive loss function value, the corresponding importance weight coefficient is temporarily increased for the loads carried by equipment at high-risk or early warning levels. When evaluating electrical isolation schemes, priority should be given to ensuring power supply continuity for loads with high weighting factors.

[0017] By adopting the above technical solutions, and synchronizing the thermal risk level or fault warning level of equipment with the decision-making logic of the self-healing process, the formulation of fault handling plans is no longer based solely on the current fault state, but rather takes into account the existing risks and hidden dangers of the equipment in advance, making the decision-making logic more forward-looking. When calculating the load loss index of the comprehensive loss function, the importance weight of the loads connected to high-risk or warning-state equipment is temporarily increased, further strengthening the protection priority of the loads associated with the risky equipment, avoiding load interruptions or fluctuations of high-risk equipment due to fault handling, thereby reducing the possibility of equipment fault escalation. When evaluating electrical isolation schemes, priority is given to ensuring the power supply continuity of high-weight loads, which can minimize the impact on critical loads during fault handling, taking into account the dual requirements of power supply reliability and equipment risk control.

[0018] Secondly, this application provides a BIM-based safety monitoring system for power electromechanical equipment, employing the following technical solution: A BIM-based safety monitoring system for power electromechanical equipment includes: The digital twin construction module is used to construct a digital twin that maps to the physical power system by integrating the three-dimensional geometric model, thermophysical property model, real-time electrical topology model, and control logic model of the power equipment. The data synchronization module is used to acquire multi-source heterogeneous data of the physical power system and synchronize the multi-source heterogeneous data to the digital twin; the multi-source heterogeneous data includes the collected equipment casing temperature, load current, environmental parameters, and the acquired protection signals and switch status; The predictive maintenance analysis module is used to solve the inverse problem of heat conduction in the digital twin at a preset first fixed period, based on the synchronized equipment casing temperature, load current, and environmental parameters, combined with the thermophysical property model, to deduce the predicted temperature field and heat source power inside the equipment; and to evaluate the thermal risk level of the equipment by comparing the predicted temperature field with the preset equipment insulation heat resistance threshold. When the warning or alarm level is reached, a predictive maintenance work order containing equipment location and predicted hotspot information is generated and output. The real-time monitoring and early warning module is used to process the multi-source heterogeneous data synchronized to the digital twin in real time and match it with the preset early fault model. If a match is found, a corresponding fault warning is generated and output. At the same time, the module monitors the protection signals and switch status in real time. When an event that meets the preset electrical fault judgment conditions is detected, the self-healing process is triggered. The autonomous healing decision module is used to determine the fault range based on the real-time electrical topology model in the digital twin after the autonomous healing process is triggered, generate multiple feasible electrical isolation schemes, and determine the optimal scheme from the multiple electrical isolation schemes based on preset evaluation criteria. The instruction execution module is used to perform safety logic verification on the optimal solution, generate and send a sequence of control instructions to the physical system to perform fault isolation and power restoration to non-faulty sections.

[0019] Thirdly, this application provides a computer device that adopts the following technical solution: A computer device includes a memory, a processor, and a computer program stored in the memory, the processor executing the computer program to implement the BIM-based safety monitoring method for power electromechanical equipment as described in the first aspect. Attached Figure Description

[0020] Figure 1 This is a first flowchart of an embodiment of the method of this application; Figure 2 This is a second flowchart of an embodiment of the method of this application; Figure 3 This is a third flowchart of an embodiment of the method of this application; Figure 4 This is the fourth flowchart of an embodiment of the method of this application; Figure 5 This is the fifth flowchart of an embodiment of the method of this application; Figure 6 This is the sixth flowchart of an embodiment of the method of this application. Detailed Implementation

[0021] To make the purpose, technical solution, and advantages of this application clearer, the following description is provided in conjunction with the appendix. Figures 1-6 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

[0022] The first embodiment of this application discloses a BIM-based safety monitoring method for power electromechanical equipment. (Refer to...) Figure 1 The method includes S110-S170: S110 constructs a digital twin that maps the physical power system by integrating the three-dimensional geometric model, thermophysical property model, real-time electrical topology model, and control logic model of the power equipment. S120 acquires multi-source heterogeneous data of the physical power system and synchronizes the multi-source heterogeneous data to the digital twin; the multi-source heterogeneous data includes the collected equipment casing temperature, load current, environmental parameters, as well as the acquired protection signals and switch status; S130, with a preset first fixed cycle, in the digital twin, based on the synchronized device shell temperature, load current and environmental parameters, combined with the thermophysical property model, the inverse problem of heat conduction is solved, and the predicted temperature field and heat source power inside the device are derived. S140 compares the predicted temperature field with the preset equipment insulation heat resistance threshold to assess the equipment thermal risk level. When the warning or alarm level is reached, a predictive maintenance work order containing equipment location and predicted hot spot information is generated and output. S150 processes multi-source heterogeneous data synchronized to the digital twin in real time and matches it with a preset early fault model. If a match is found, a corresponding fault warning is generated and output. At the same time, it monitors protection signals and switch status in real time. When an event that meets the preset electrical fault judgment conditions is detected, the autonomous healing process is triggered. S160, after the autonomous healing process is triggered, the fault area is determined based on the real-time electrical topology model in the digital twin, and multiple feasible electrical isolation schemes are generated. The optimal scheme is determined from the multiple electrical isolation schemes based on the preset evaluation criteria. After safety logic verification, S170 generates and sends a sequence of control commands to the physical system to perform fault isolation and restore power supply to non-faulty sections.

[0023] Specifically, for step S110, core BIM modeling software (such as Revit and Bentley) is used as the basic modeling tool. First, detailed CAD design drawings of the electrical equipment (such as switchgear, transformers, and cable trays) are imported to construct a 3D geometric model of the equipment. The dimensions, locations, and spatial relationships of key components such as the equipment casing, internal components, and terminals are refined to ensure a 1:1 correspondence between the 3D model and the physical equipment. Thermophysical property parameters of the equipment are collected, including the thermal conductivity and specific heat capacity of the casing material (such as cold-rolled steel plate and aluminum alloy), the thermal conductivity characteristics of the internal insulation material (such as Class E and Class B insulation), and the thermal resistance parameters of the conductors (copper busbars and aluminum busbars). A thermophysical property model is established using thermal simulation software such as ANSYS and Fluent, and the thermophysical parameters are linked and bound to the BIM 3D geometric model to visualize the model's thermal characteristics. For the real-time electrical topology model, based on the primary wiring diagram of the power system, the electrical connection relationships between devices are constructed in the BIM model, including the wiring logic of devices such as busbars, circuit breakers, and disconnect switches. A graph database is used to store the topology node information to ensure that the topology relationships can be updated in real time. The control logic model replicates the PLC control program and protection device logic of the devices, embedding information such as switch action logic, protection settings, and control circuits into the BIM model. Finally, through a unified data interface (such as the IFC standard interface), the three-dimensional geometric model, thermophysical property model, real-time electrical topology model, and control logic model are integrated to construct a complete digital twin, achieving accurate mapping of the physical power system.

[0024] For step S120, the acquisition of multi-source heterogeneous data adopts a dual approach of "sensor acquisition + system integration". For equipment casing temperature, infrared temperature sensors and contact temperature sensors are installed at key heat-generating areas of the equipment casing (such as busbar compartments, circuit breaker compartments, and cable joints). The sensor sampling frequency is set to 1 time / minute, and the sampling accuracy is controlled within ±0.5℃. The temperature data is transmitted in real time to the edge computing gateway via a wireless LoRa module. Load current data is acquired through Rogowski coil current sensors and Hall effect current sensors, installed at the equipment's inlet and outlet terminals, collecting the effective value of the three-phase current at a sampling frequency of 5 times / second to ensure the capture of dynamic current changes. Environmental parameters (temperature, humidity, dust concentration) are acquired through temperature and humidity sensors and dust sensors deployed in the equipment room and around the equipment, with a sampling frequency of 1 time / 5 minutes. Protection signals and switch status are obtained by integrating with the substation SCADA system and the equipment PLC control system, using the OPC UA communication protocol to read overcurrent protection, grounding protection, gas protection, and other protection action signals, as well as switch opening / closing status and energy storage status information in real time. When synchronizing data to the digital twin, the collected data is first preprocessed through the edge computing gateway. The Kalman filter algorithm is used to remove outliers in the temperature and current data (such as temperature values ​​that deviate from the normal range of ±3℃ due to sensor failure, and sudden changes in current values). Missing data is supplemented by linear interpolation. Then, the preprocessed data is synchronized to the digital twin through the MQTT protocol, with the data synchronization delay set to no more than 500ms to ensure data consistency between the digital twin and the physical power system.

[0025] Reference Figure 2 In S130, the steps of solving the inverse problem of heat conduction by combining the thermophysical property model to deduce the predicted temperature field and heat source power inside the equipment specifically include S210-S240: S210, calculates the average temperature of the housing based on the readings of multiple temperature sensors arranged on the housing of the equipment; S220, based on the collected effective value of the load current, estimate the reference heating power of the equipment; S230, establish a heat conduction equation with the internal hot spot temperature of the equipment as the unknown core and the average temperature of the outer shell and the ambient temperature as boundary conditions; S240 uses an optimization algorithm to iteratively adjust the assumed value of the internal heat source power, so that the error between the shell temperature distribution calculated based on the assumed value and the measured temperature distribution is minimized. At this time, the corresponding heat source power and the calculated internal temperature are the inversion results.

[0026] Specifically, for step S210, the preset first fixed cycle is set to 5 minutes based on the equipment's operating characteristics. This cycle ensures real-time monitoring while avoiding data redundancy and computational pressure caused by excessively short cycles. In the digital twin, readings from multiple temperature sensors located on the synchronized equipment casing are retrieved. First, the sensor data is validated, and invalid data (such as data from offline sensors or data with abnormal fluctuations) is removed. If the amount of valid data is less than 80% of the total number of sensors, valid temperature data from the corresponding position in the previous cycle is extracted, and the missing sensor data in the current cycle is supplemented using the Lagrange interpolation algorithm. Then, the arithmetic mean temperature is calculated. If the amount of valid data meets the requirements, the arithmetic mean method is used to calculate the average casing temperature. The formula is: Average casing temperature = (T1 + T2 + ... + Tn) / n, where T1 to Tn are the readings of each valid temperature sensor, and n is the number of valid sensors. Simultaneously, the calculated average casing temperature and corresponding timestamp are recorded in the digital twin to form a temperature change trend curve, providing a basis for subsequent heat conduction analysis and temperature anomaly judgment.

[0027] For step S220, the conductor parameters of the corresponding device are retrieved from the device attribute library of the digital twin, including conductor material (copper, aluminum), cross-sectional area, length, etc., and the resistivity at 20°C is determined based on the conductor material. (For example, the resistivity of copper is 1.72 × 10⁻⁶) -8 Ω·m), combined with the conductor length L and cross-sectional area S, through the formula The resistance of the conductor at 20℃ was calculated. Then, considering the effect of temperature on the conductor's resistance, a temperature correction factor was introduced. (The temperature correction factor for copper is 0.004 / ℃), using the formula Calculate the actual conductor resistance at the current average casing temperature Ts. Based on the synchronously acquired effective load current I, calculate the single-phase reference heating power P1=I using Joule's law. 2 If Rt is the reference heating power of the equipment, then P = 3 × P1 (for a three-phase balanced system), or the three-phase currents can be substituted and summed. This power reflects the basic heating level of the equipment under the current load, providing a basic heat source reference for subsequent internal hot spot temperature inversion.

[0028] For step S230, considering the structural characteristics of the equipment (such as the compartmentalized structure of the switchgear and the core structure of the transformer), the equipment is simplified into a three-dimensional steady-state heat conduction model. A heat conduction equation is established with the internal hot spot temperature as the unknown core and the average shell temperature and ambient temperature as boundary conditions. The heat conduction equation is derived using Fourier's law, and the expression for the three-dimensional steady-state heat conduction equation is ∇⋅(k∇T)+q=0, where k is the thermal conductivity of the internal materials (obtained from the thermophysical property model; different materials use corresponding thermal conductivity, such as copper with a thermal conductivity of 380 W / (m·℃) and insulating materials with a thermal conductivity of 0.2 W / (m·℃)), T is the temperature field distribution, and q is the internal heat source power density (unit: W / m³). 3 The relationship between q and the baseline heating power P estimated in S220 is q = P / V, where V is the equivalent volume of the heating conductor. During the iteration process, the q value is updated by adjusting P, and ∇ is the Laplace operator. Boundary conditions are divided into two categories: the first category is the surface temperature of the equipment casing, i.e., the average casing temperature Ts; the second category is the convective heat transfer boundary between the equipment casing and the environment, described using Newton's law of cooling, i.e. Where h is the convective heat transfer coefficient (set according to the ventilation conditions of the computer room; h is 15-25 W / (m²·℃) for natural ventilation and 30-50 W / (m²·℃) for forced ventilation), Ta is the ambient temperature, and n is the normal direction of the equipment shell. By substituting the boundary conditions into the heat conduction equation, a complete heat conduction model is constructed, providing a mathematical basis for subsequent internal hot spot temperature inversion.

[0029] For step S240, Particle Swarm Optimization (PSO) is used as the optimization algorithm. This algorithm features fast convergence and strong global search capability, making it suitable for iterative adjustment of internal heat source power. First, the search range for internal heat source power is set. Based on the estimated baseline heating power, the search range is set to 0.8-1.2 times the baseline heating power to ensure the reasonableness of the search range and avoid deviations from actual assumed values. Then, the particle swarm is initialized, with each particle representing an assumed value for internal heat source power. The number of particles is set to 50, the maximum number of iterations is set to 100, and the convergence threshold is set to 0.25℃. 2During the iteration process, each particle updates its velocity and position based on its fitness value (i.e., the sum of squared errors between the calculated shell temperature distribution and the measured temperature distribution). The smaller the fitness value, the closer the assumed heat source power is to the actual value. Specifically, for each particle, the assumed heat source power is substituted into the established heat conduction equation to calculate the temperature distribution of the equipment shell. This is compared with the collected measured temperature distribution, and the sum of squared errors is calculated as the fitness value. The particle's velocity and position are iteratively updated to gradually reduce the error. Iteration stops when the fitness value is less than the convergence threshold. The corresponding heat source power at this point is the actual internal heat source power. The internal temperature calculated using the heat conduction equation is the inversion result (internal hot spot temperature). For example, if the reference heating power of a switchgear is 527.04W, after initializing the particle swarm, through 100 iterations, when the error drops to 0.3℃, the corresponding heat source power is 580W, and the calculated internal hot spot temperature is 95℃. This result is the inverted actual internal heating of the equipment.

[0030] Reference Figure 3 In S140, the steps for assessing the thermal risk level of equipment and generating predictive maintenance work orders specifically include S310-S350: S310, calculate the difference between the inverted internal hot spot temperature and the maximum allowable temperature of the equipment insulation material to obtain the temperature rise margin; S320, calculates the rate of change of the internal hot spot temperature between the current cycle and the previous cycle to obtain the temperature rise rate; S330, when the temperature rise margin is lower than the first threshold or the temperature rise rate is positive and exceeds the set rate, it is determined to be a warning level; S340, when the temperature rise margin is lower than the lower second threshold or the internal hot spot temperature reaches the allowable limit, it is determined to be an alarm level; In the S350, the corresponding equipment model is highlighted in color according to the risk level in the three-dimensional geometric model of the digital twin, and a predictive maintenance work order is automatically generated, which includes the equipment's unique identifier, predicted hot spot locations, suggested maintenance items, and recommended time windows.

[0031] Specifically, for step S310, the maximum allowable temperature of the corresponding equipment insulation material is retrieved from the equipment attribute library of the digital twin. Different insulation grades correspond to different maximum allowable temperatures (e.g., Class E insulation has an allowable temperature of 120℃, Class B insulation has 130℃, and Class F insulation has 155℃). If different parts of the equipment use different grades of insulation materials, the maximum allowable temperature of each part is retrieved separately, and the minimum value is taken as the maximum allowable temperature of the entire equipment to ensure the safety of monitoring. Based on the internal hot spot temperature obtained through inversion, the temperature rise margin is calculated. The formula for calculating the temperature rise margin is: Temperature rise margin = Maximum allowable temperature of insulation material - Inverted internal hot spot temperature. The temperature rise margin reflects the current thermal safety margin of the equipment. The larger the value, the better the thermal safety performance of the equipment; conversely, the smaller the value, the closer it is to the thermal limit. For example, a switchgear using Class E insulation material has a maximum allowable temperature of 120℃. If the inverted internal hot spot temperature is 95℃, then the temperature rise margin = 120 - 95 = 25℃, indicating that the equipment currently has a 25℃ thermal safety margin and is operating well. If the inverted internal hot spot temperature is 110℃, then the temperature rise margin = 10℃, indicating that the equipment is approaching its thermal limit and requires close monitoring. The calculated temperature rise margin is synchronized to the digital twin in real time.

[0032] For step S320, based on the set first fixed period (5 minutes), the internal hotspot temperature (Tcurrent) retrieved from the current period and the internal hotspot temperature retrieved from the previous period (Tprevious period) are extracted. If data is missing from the previous period, linear interpolation is used to supplement the internal hotspot temperature. The temperature rise rate is calculated using the formula: Temperature rise rate = (Tcurrent - Tprevious period) / Tperiod, where Tperiod is the first fixed period (5 minutes), and the unit is °C / minute. The temperature rise rate reflects the changing trend of the internal hotspot temperature of the equipment; a positive value indicates that the temperature is rising, and a negative value indicates that the temperature is falling. The larger the absolute value, the faster the temperature changes. For example, if the current internal hotspot temperature is 95℃ and the previous cycle was 93℃, with a cycle duration of 5 minutes, then the temperature rise rate is (95-93) / 5 = 0.4℃ / minute, indicating that the internal temperature of the equipment is rising at a rate of 0.4℃ / minute. It is necessary to monitor the temperature trend and prevent a rapid temperature increase. If the current cycle temperature is 92℃ and the previous cycle was 95℃, then the temperature rise rate is (92-95) / 5 = -0.6℃ / minute, indicating that the temperature is decreasing and the thermal condition of the equipment is improving. Simultaneously, the calculated temperature rise rate for each cycle is recorded to form a temperature rise rate change curve, which, combined with the temperature rise margin, provides a basis for risk level determination.

[0033] For step S330, a first threshold and a set rate are preset. The first threshold is set based on the equipment's operating years, maintenance records, and insulation material characteristics. For equipment with a short operating life (≤5 years) and timely maintenance, the first threshold is set to 15℃; for equipment with a long operating life (>5 years) and minor defects, the first threshold is set to 12℃, ensuring the reasonableness and relevance of the threshold setting. The set rate is set to 0.3℃ / minute based on the equipment's thermal characteristics. This rate is the safe critical value for the equipment's temperature rise; exceeding this rate indicates that the temperature rise is too rapid and there is a thermal risk. During the judgment process, the temperature rise margin and temperature rise rate are retrieved in real time. If one of the following two conditions is met, it is judged to be at the warning level: First, the temperature rise margin is lower than the preset first threshold (e.g., 15℃). For example, if the temperature rise margin is 12℃, which is lower than 15℃, it indicates that the equipment's thermal safety margin is insufficient and there is a potential thermal risk. Second, the temperature rise rate is positive and exceeds the set rate (0.3℃ / minute). For example, if the temperature rise rate is 0.4℃ / minute and is positive, it indicates that the equipment temperature is rising rapidly, and if not intervened in time, the temperature may exceed the allowable limit. After being judged to be at the warning level, the warning status of the equipment is marked in the digital twin, and the warning information is pushed to the monitoring platform simultaneously to remind maintenance personnel to pay attention to the equipment's operating status and prepare for troubleshooting.

[0034] For step S340, a second threshold (lower than the first threshold) and the maximum allowable temperature of the equipment insulation material are preset. The second threshold is set based on the first threshold, usually 2 / 3 of the first threshold. For example, if the first threshold is 15℃, the second threshold is set to 10℃; if the first threshold is 12℃, the second threshold is set to 8℃, ensuring that the alarm level determination is more stringent than the warning level. The maximum allowable temperature of the equipment insulation material serves as the absolute safety limit for the internal hot spot temperature. During the determination process, the calculated temperature rise margin and the inverted internal hot spot temperature are retrieved in real time. If one of the following two conditions is met, an alarm level is determined: First, the temperature rise margin is lower than the preset second threshold (e.g., 10℃). For example, if the temperature rise margin is 8℃ and lower than 10℃, it indicates that the equipment's thermal safety margin is severely insufficient and may reach the thermal limit at any time; second, the internal hot spot temperature reaches the maximum allowable temperature of the insulation material (e.g., 120℃ for Class E insulation). At this point, the equipment is in a thermal hazard state, and if not dealt with immediately, it may lead to insulation aging, equipment failure, or even fire. Once an alarm is detected, an emergency response procedure is immediately triggered. The device is marked as alarmed in the digital twin, and alarm information is simultaneously pushed to the mobile terminals of maintenance personnel and the monitoring platform, requiring maintenance personnel to rush to the site for investigation and handling.

[0035] For step S350, in the three-dimensional geometric model of the digital twin, the rendering function of BIM software is used to set corresponding color highlighting rules according to the risk level of the equipment (normal, warning, alarm): the normal state is rendered as green, the warning state as yellow, and the alarm state as red. The highlighting range covers the entire equipment model. At the same time, the predicted hot spot locations (such as "the upper joint of the busbar compartment") are marked on the model to facilitate the operation and maintenance personnel to quickly locate the risk areas. When generating predictive maintenance work orders, the system first retrieves the unique identifier of the equipment (e.g., equipment number KYN28-12-001) from the digital twin. This identifier corresponds one-to-one with the physical equipment, ensuring the accuracy of the work order. Then, combining the inverted predicted hotspot locations, temperature rise margin, temperature rise rate, and historical maintenance records of the equipment, it generates suggested maintenance items, such as "check for loose busbar joints and tighten the connecting bolts; clean the dust inside the equipment and improve heat dissipation." The recommended time window is set according to the risk level: equipment at the warning level is recommended to be maintained within 72 hours, and equipment at the alarm level is recommended to be maintained within 24 hours, ensuring timely maintenance. The work order format uses a standardized template, including the work order number, unique equipment identifier, predicted hotspot location, risk level, suggested maintenance items, recommended maintenance time, and responsible maintenance personnel. After generation, it is automatically synchronized to the maintenance management system and pushed to the corresponding maintenance personnel's terminals, facilitating the scheduling of maintenance work.

[0036] For step S150, multi-source heterogeneous data synchronized to the digital twin is processed in real time. An edge computing platform is used to extract features from the data, such as the harmonic distortion rate of current data, the slope of temperature data change, and the fluctuation range of environmental parameters. The extracted features are matched with a preset early fault model. The early fault model is built based on historical fault data and includes feature parameters of common faults such as loose bus joints, insulation aging, and poor contact (e.g., when the bus joint is loose, the harmonic distortion rate of current exceeds 5% and the temperature rise rate exceeds 0.5℃ / minute). If the feature matching degree exceeds 80%, it is determined that an early fault exists, and a corresponding fault warning is generated and output, clarifying the fault type, suspected fault location, and preliminary handling suggestions, which are then pushed to the monitoring platform simultaneously.

[0037] Meanwhile, the protection signals and switch status are monitored in real time through the OPC UA protocol, and electrical fault judgment conditions are preset, such as overcurrent protection activation and switch tripping, grounding protection activation and fault current exceeding 1.5 times the rated current, gas protection activation and abnormal oil level, etc. When an event that meets the judgment conditions is detected, the self-healing process is immediately triggered to buy time for fault handling.

[0038] The steps for determining the optimal solution from multiple electrical isolation schemes based on preset evaluation criteria specifically include: Centered on the fault point, all feasible electrical isolation schemes are simulated in parallel within a digital twin. For each electrical isolation scheme, online simulation is performed and the comprehensive loss function value is calculated. The electrical isolation scheme with the smallest comprehensive loss function value is selected as the optimal scheme.

[0039] Reference Figure 4 The steps for performing online simulation and calculating the comprehensive loss function value for each electrical isolation scheme specifically include S410-S460: S410 simulates the switching operation of the corresponding electrical isolation scheme in the digital twin copy and updates the electrical topology connection state; S420 performs fast power flow estimation based on the updated electrical topology to obtain the reconstructed current values ​​of each branch and the voltage values ​​of each node. S430, based on the reconstructed current value and combined with the equipment thermal model, quickly estimates the new heat generation power and hot spot temperature of key equipment, and calculates potential thermal risk indicators; S440: Statistically calculate the load loss index by analyzing the power loss load caused by the corresponding electrical isolation scheme and combining it with the preset load importance weight. S450: Calculate the number of switches to be operated and the physical distance of the operation path required for the corresponding electrical isolation scheme, and calculate the operation complexity index. S460 calculates the comprehensive loss function value of the corresponding electrical isolation scheme based on the load loss index, operation complexity index, and potential thermal risk index using a weighted summation algorithm.

[0040] In addition, after generating predictive maintenance work orders or fault warnings, the corresponding thermal risk level (warning / alarm) or fault warning level (general / emergency) of the equipment is synchronized to the decision logic of the self-healing process. For example, for equipment faults at the alarm level, when selecting an electrical isolation scheme, priority is given to ensuring the power supply continuity of the important loads connected to the equipment, so as to avoid greater losses due to isolation operations.

[0041] Specifically, for step S410, after the autonomous healing process is triggered, a depth-first search (DFS) algorithm is used to determine the fault range based on the real-time electrical topology model in the digital twin. This algorithm can quickly traverse the electrical topology nodes and locate the fault range based on fault signals (such as protection action signals and switch tripping signals). For example, if a short-circuit fault occurs on a feeder, the fault range can be located as the line and equipment between the feeder switch and the load switch by traversing the topology nodes. Subsequently, based on the location of the fault range, the electrical topology structure, and the equipment distribution, multiple feasible electrical isolation schemes are generated. The generation of isolation schemes must comply with electrical safety regulations to ensure that the fault range is completely disconnected from the system after isolation and does not affect the normal power supply of non-faulty sections. For example, for the aforementioned feeder fault, two feasible schemes can be generated: Scheme 1 is to disconnect the feeder switch and the load switch to isolate the fault range; Scheme 2 is to disconnect the feeder switch and the sectionalizing switch to isolate the fault range.

[0042] After generating feasible solutions, all feasible solutions are simulated in parallel in the digital twin, centered on the fault point. The electrical topology changes after the isolation solution is executed are simulated to check the feasibility of the solution (such as whether there are problems such as power loss or overvoltage in non-faulty sections). Infeasible solutions are eliminated to ensure that the retained solutions can safely and effectively isolate the faulty section.

[0043] The switching operations of electrical isolation schemes are simulated in a digital twin replica. The digital twin replica is synchronized with the master model in real time to ensure that the simulation environment is consistent with the actual operating environment. For each feasible electrical isolation scheme, the switching operation sequence is set according to electrical operation specifications (e.g., disconnect the load switch first, then disconnect the power switch; close the power switch first, then close the load switch). The opening and closing operations of the corresponding switches are simulated in the digital twin replica. For example, the operation sequence for Scheme 1 is: first disconnect the load switch K1, wait 5 seconds to confirm that the opening is in place, and then disconnect the feeder switch K2. During the simulation operation, the electrical topology connection status of the digital twin replica is updated in real time. Faulty sections are marked as isolated, while non-faulty sections remain in a normal connection state. At the same time, the operation time and operation status (opening / closing) of each switch are recorded to provide data support for subsequent scheme evaluation.

[0044] For step S420, a simplified version of the Newton-Raphson method is used for fast power flow estimation based on the updated electrical topology connection status. This algorithm is fast and accurate, meeting the real-time requirements of the self-healing process (calculation time controlled within 10 seconds). First, system power parameters are input, including bus rated voltage, generator output, and load power, retrieved from the real-time database of the digital twin. Then, based on the updated electrical topology, a node admittance matrix is ​​established, and the voltage of each node and the current of each branch are calculated iteratively to obtain the reconstructed current values ​​of each branch and the voltage values ​​of each node. For example, after fault isolation, the reconstructed current of a certain branch drops from 1200A to 800A, and the node voltage stabilizes at 10.3kV from 10.5kV, both within the allowable range (voltage deviation not exceeding ±5%), indicating that under this isolation scheme, the electrical parameters of the non-faulty section are stable, and the power supply quality can be guaranteed. Simultaneously, the power flow estimation results are fed back to the digital twin replica in real time, providing data support for subsequent thermal risk assessment and scheme optimization.

[0045] For step S430, based on the reconstructed current value and the equipment thermal model established in S140, the new heat generation power of critical equipment (such as busbars, circuit breakers, transformers, etc.) is quickly estimated. The estimation method is consistent with S140, that is, the Joule heat power is calculated through the effective value of the current and the actual resistance of the conductor. At the same time, the impact of the change in current after reconstruction on the resistance is considered, and the resistance value is corrected to ensure the accuracy of the heat generation power estimation. Subsequently, the new heat generation power is used as a heat source and substituted into the heat conduction equation to quickly estimate the new internal hot spot temperature of the critical equipment and calculate the potential thermal risk index. The formula for calculating the potential thermal risk index is: Potential thermal risk index = 1 - (maximum allowable temperature of insulation material - new internal hot spot temperature) / maximum allowable temperature of insulation material. This index is used to assess the thermal safety status of critical equipment after the isolation scheme is implemented. The higher the index value, the higher the potential thermal risk, and vice versa.

[0046] For step S440, the electrical topology model of the digital twin is used to statistically analyze the power loss load caused by each electrical isolation scheme, including the number, power, and type of the power loss load (critical load, ordinary load). The power loss load statistics must be combined with the isolation range of the isolation scheme to ensure that no power loss load caused by isolation operations is missed. Subsequently, load importance weights are preset, setting different weight coefficients according to the importance of the load: the weight coefficient for critical loads (such as hospitals, data centers, emergency lighting, etc.) is set to 1.0, and the weight coefficient for ordinary loads (such as residential electricity, general office electricity, etc.) is set to 0.5. The weight coefficient settings can be adjusted according to user needs and power supply contracts to ensure priority power supply to critical loads. The load loss index is calculated using the formula: Load Loss Index = Σ (Power Loss Load Power × Corresponding Load Importance Weight). The smaller the index value, the smaller the load loss caused by the isolation scheme, and the better the scheme. When calculating the load importance weight, if the equipment to which the load belongs is currently in a warning or alarm thermal risk level, its weight coefficient is dynamically increased.

[0047] For step S450, the number of switches required for each electrical isolation scheme is counted, including the total number of opening and closing operations. A higher number of switches indicates more complex and cumbersome operation steps. Based on the 3D geometric model of the digital twin, the physical distance of the operator's path is calculated. The path starts from the control room, passes through each switch requiring operation, and the path length is calculated using the 3D coordinates of the BIM model according to the actual walking route. A longer path distance indicates longer operation time and higher complexity. An operation complexity index is calculated using the formula: Operation Complexity Index = (Number of Switch Operations × Standard Time for a Single Switch Operation) + (Path Distance / Average Personnel Movement Speed). This formula quantifies the number of switches and path distance to objectively assess operation complexity; a smaller index value indicates simpler and more efficient operation.

[0048] For step S460, weighting coefficients for the comprehensive loss function are set. Based on the priorities of power supply reliability, operational efficiency, and equipment safety, the weights are set as follows: load loss index w1 = 0.5, operational complexity index w2 = 0.3, and potential thermal risk index w3 = 0.2. The sum of these three weighting coefficients is 1, and they can be fine-tuned according to actual operational needs. Then, the calculated load loss index, operational complexity index, and potential thermal risk index are normalized. Finally, the normalized data is used to calculate the comprehensive loss function value for each electrical isolation scheme using a weighted summation algorithm. The calculation formula is: Comprehensive loss function value = w1 × load loss index + w2 × operational complexity index + w3 × potential thermal risk index. The electrical isolation scheme with the smallest comprehensive loss function value is the optimal scheme. Furthermore, when calculating load loss indicators, the corresponding importance weighting coefficient is temporarily increased for loads supplied by equipment at high-risk (alarm) or early warning levels. For example, the weighting coefficient for ordinary loads supplied by early warning equipment is increased from 0.5 to 0.7, and the weighting coefficient for ordinary loads supplied by alarm equipment is increased to 0.9. This ensures priority is given to ensuring the continuity of power supply to loads supplied by high-risk equipment, preventing further exacerbation of equipment thermal risks due to load power failure. When evaluating electrical isolation schemes, if the comprehensive loss function values ​​of two schemes are close, the scheme that can guarantee the power supply to loads with higher weighting coefficients is preferred. For example, if scheme A has a comprehensive loss of 38 and scheme B has a comprehensive loss of 39, but scheme A can guarantee the power supply to important loads supplied by alarm equipment while scheme B cannot, then scheme A is preferred.

[0049] For step S170, a safety logic verification is performed on the control command sequence corresponding to the optimal electrical isolation scheme. The verification includes whether the switch operation sequence complies with electrical safety regulations (e.g., whether there are violations such as disconnecting the power supply before disconnecting the load, or closing the load before closing the power supply), whether the control commands will cause overvoltage or overcurrent in non-faulty sections, and the completeness and accuracy of the commands (e.g., whether the switch number and operation time are correct). The verification uses logic simulation, simulating the execution of the control command sequence in a digital twin copy, checking for any abnormalities during execution. If an abnormality occurs, the command sequence is immediately adjusted until the verification passes. After successful verification, a control command sequence is generated based on the optimal electrical isolation scheme. The command sequence includes information such as switch operation number, operation type (opening / closing), operation time, and operation sequence, for example, "10:00:00 Open load switch K1, 10:00:05 Open feeder switch K2, 10:00:10 Close standby switch K3." Subsequently, the control command sequence is sent to the control devices such as PLCs and RTUs of the physical system via the IEC61850 communication protocol to perform fault isolation and power restoration operations in non-faulty sections. After the commands are sent, the operation feedback of the physical system is monitored in real time. If a switch operation fails (such as a switch refusing to operate), a backup control command sequence is immediately generated, and the operation plan is adjusted to ensure the smooth completion of fault isolation and power restoration, thus guaranteeing the stable operation of the power system.

[0050] Reference Figure 5 Furthermore, the safety monitoring methods also include S510-S530: S510: After the equipment completes maintenance according to the predictive maintenance work order, obtain the actual location and severity information of the defects in the maintenance record; S520 compares the real defect information with the hotspot locations and temperatures predicted by the digital twin inversion and calculates the prediction deviation; S530 adjusts the corresponding parameters of the thermophysical property model in the digital twin in reverse according to the prediction deviation, so as to reduce the gap between the virtual model and the physical entity.

[0051] Specifically, for step S510, after the equipment completes maintenance according to the predictive maintenance work order, maintenance personnel enter detailed maintenance records through the maintenance management system. These records must include maintenance time, maintenance personnel, unique equipment identifier, actual defect location, defect severity, maintenance measures, and maintenance results. The actual defect location must be accurately described in conjunction with the equipment's three-dimensional structure (e.g., "Loose copper busbar connector on the left side of the switchgear busbar compartment, contact resistance increased by 10%"). Defect severity is categorized into three levels: mild, moderate, and severe. Mild defects do not affect normal equipment operation and only require periodic monitoring; moderate defects affect equipment operating efficiency and require immediate attention; severe defects threaten safe equipment operation and require immediate maintenance. Simultaneously, during maintenance, personnel use infrared thermometers, megohmmeters, and other equipment to measure the actual temperature and insulation resistance of the equipment after maintenance and enter this data into the maintenance records as a basis for subsequent model calibration. The system automatically extracts the actual defect location and severity information from the maintenance management system and associates it with the corresponding digital twin equipment.

[0052] For step S520, firstly, the hotspot location and temperature data predicted before maintenance are retrieved from the digital twin and compared with the actual defect location in the maintenance record and the actual temperature data measured after maintenance. The location comparison uses a three-dimensional coordinate comparison to calculate the straight-line distance between the predicted hotspot location and the actual defect location, which is taken as the location deviation (e.g., if the predicted hotspot location is on the right side of the busbar compartment and the actual defect location is on the left side, the deviation distance is 50mm). The temperature comparison calculates the difference between the predicted internal hotspot temperature and the actual internal hotspot temperature measured after maintenance, which is taken as the temperature deviation (e.g., if the predicted temperature is 95℃ and the actual measured temperature is 75℃, the temperature deviation is 20℃). Then, the comprehensive prediction deviation is calculated. The formula for the comprehensive prediction deviation is: Comprehensive prediction deviation = (Location deviation / Corresponding part size of equipment × 50%) + (Temperature deviation / Maximum allowable temperature of insulation material × 50%). This formula takes into account both location and temperature deviations and can comprehensively reflect the prediction accuracy of the digital twin. For example, if the busbar compartment of the equipment is 1000mm in size, the positional deviation is 50mm, the temperature deviation is 20℃, and the maximum allowable temperature of the insulation material is 120℃, then the comprehensive prediction deviation = (50 / 1000×50%) + (20 / 120×50%) ≈ 2.5% + 8.33% = 10.83%. The larger the comprehensive prediction deviation, the greater the difference between the digital twin and the physical entity, and parameter adjustments are required.

[0053] For step S530, based on the calculated comprehensive prediction deviation, the corresponding parameters of the thermophysical property model in the digital twin are adjusted in reverse. The adjustment principle is as follows: If the temperature deviation is large (accounting for more than 60% of the comprehensive prediction deviation), the focus is on adjusting parameters such as the thermal conductivity and convective heat transfer coefficient of the internal materials of the equipment. For example, if the temperature deviation is large and the predicted temperature is higher than the actual temperature, it indicates that the thermal conductivity in the model is set too low, and the thermal conductivity needs to be appropriately increased (e.g., combined with the actual contact thermal resistance, the equivalent thermal conductivity is adjusted from 350 W / (m·℃) to 380 W / (m·℃)) to make the temperature calculated by the model closer to the actual temperature. If the position deviation is large (accounting for more than 60% of the comprehensive prediction deviation), the mapping coordinates of the temperature sensor in the digital twin are adjusted to ensure that the sensor data corresponds to the actual part of the equipment. At the same time, the boundary conditions of the heat conduction equation are adjusted to optimize the calculation accuracy of the temperature field distribution. After the adjustment is completed, the internal hot spot temperature of the equipment is re-inverted, and the new comprehensive prediction deviation is calculated. If the deviation is still higher than the preset threshold (e.g., 5%), the iterative adjustment continues until the comprehensive prediction deviation is lower than the threshold, thus narrowing the gap between the virtual model and the physical entity.

[0054] Reference Figure 6 Furthermore, the safety monitoring methods also include S610-S630: S610 records the actual handling results of each fault event, including the actual isolation range, load recovery time, and whether it caused subsequent problems; S620 compares and analyzes the actual handling results with the comprehensive loss function values ​​of each candidate electrical isolation scheme during online simulation in a digital twin; S630: When the actual optimal solution is inconsistent with the solution selected in the simulation, or when there is a systematic deviation between the simulation evaluation index and the actual disposal result, the weight coefficients of load loss, operation complexity and potential thermal risk in the comprehensive loss function are iteratively adjusted.

[0055] Specifically, for step S610, a fault event database is established to record the actual handling results of each fault event in real time. The records include the fault occurrence time, fault type, fault location, actual electrical isolation scheme used, actual isolation range, load restoration time, problems encountered during the handling process and their solutions, and whether subsequent faults (such as secondary short circuits or equipment damage) were triggered. For example, for a feeder short circuit fault, Scheme 1 was actually used for isolation, with the isolation range from the feeder switch to the load switch. The load restoration time was 20 minutes. No problems such as switch failure to operate occurred during the handling process, and no subsequent faults were triggered. This information is completely entered into the fault event database and simultaneously linked to the corresponding digital twin simulation data for subsequent comparative analysis. For problems encountered during the handling process (such as switch failure to operate or load restoration delays), the causes and solutions are recorded in detail to provide a basis for subsequent optimization of the simulation model and decision-making logic.

[0056] For step S620, the actual handling results of a specific fault event are extracted from the fault event database, including information such as the actual isolation range, load recovery time, load loss, and operation time. This data is then compared with the comprehensive loss function value, simulated isolation range, simulated load recovery time, and simulated load loss of each candidate electrical isolation scheme in the online simulation of the same fault event in the digital twin. The comparative analysis focuses on: whether the actual optimal scheme is consistent with the optimal scheme selected in the simulation; the deviation between the simulated predicted load loss and the actual load loss; the deviation between the simulated predicted operation time and the actual operation time; and the deviation between the simulated predicted potential thermal risks and the actual thermal state of the equipment. For example, in the simulation, the optimal scheme is Scheme 1, with a comprehensive loss function value of 38 and a predicted load recovery time of 18 minutes. In the actual implementation of Scheme 1, the load recovery time is 20 minutes (a deviation of 2 minutes), the simulated load loss is 50, and the actual load loss is 52 (a deviation of 4%). Through comparative analysis, the degree of deviation between the simulation model and the actual situation is clarified, and the causes of the deviation are identified (such as the simulation not considering the actual time of switching operations, load fluctuations, etc.).

[0057] For step S630, based on the comparative analysis results of S620 and combined with the reasons for the deviations found in the comparative analysis (such as the simulation not considering the actual time consumption of switching operations, load fluctuations, etc.), if the actual optimal solution is inconsistent with the solution selected in the simulation, or if there is a systematic deviation between the simulation evaluation indicators (such as load loss, operation complexity, potential thermal risk) and the actual handling results (such as the simulated load loss being more than 30% lower than the actual value for several consecutive times), then the weight coefficients of load loss, operation complexity and potential thermal risk in the comprehensive loss function will be iteratively adjusted. Before making adjustments, it is necessary to further clarify the root cause of the deviation. If the deviation is mainly due to inaccurate load loss assessment (such as the simulation not fully considering load fluctuations, temporary adjustments to important loads, etc.), then the weight coefficient w1 of the load loss index should be adjusted. If the actual load loss is greater than the simulated value, it indicates that w1 is set too low and needs to be increased appropriately. Each adjustment should be controlled between 0.05 and 0.1, while the weights of the other two indices should be reduced accordingly to ensure that the sum of the three weights is always 1. If the deviation is mainly due to operational complexity assessment deviation (such as the simulation not considering the actual time of switching operations, on-site operational space limitations, etc.), then the weight coefficient w2 of the operational complexity index should be adjusted. When the actual operation time is longer than the simulated value, increase w2 and decrease the weights of w1 and w3. If the deviation is mainly due to inaccurate assessment of potential thermal risks (such as the simulation not considering changes in equipment thermal characteristics after maintenance, sudden changes in ambient temperature, etc.), then the weight coefficient w3 of the potential thermal risk index should be adjusted. When the actual thermal risk is higher than the simulated value, increase w3 and decrease the weights of the other two indices. For example, if the original weighting coefficients are w1=0.5, w2=0.3, and w3=0.2, a comparison reveals that the simulated load loss was 35% lower than the actual value three consecutive times, indicating that the weight of the load loss was too low. The coefficients need to be adjusted to w1=0.6, w2=0.25, and w3=0.15. After adjustment, the parameters of the fault event are re-input into the digital twin for simulation. The new simulation results are compared with the actual handling results. If the deviation still exceeds the acceptable range (e.g., 10%), iterative adjustments continue until the deviation between the simulation evaluation index and the actual handling results is controlled within a reasonable range. Simultaneously, the adjusted weighting coefficients are updated in the decision logic of the digital twin and applied to subsequent electrical isolation scheme evaluations, ensuring the continuous improvement of the simulation model's decision-making accuracy.

[0058] Based on the above method embodiments, the second embodiment of this application discloses a BIM-based safety monitoring system for power electromechanical equipment. The BIM-based safety monitoring system for power electromechanical equipment of this embodiment can implement any of the above-described BIM-based safety monitoring methods for power electromechanical equipment, and the specific working process of each module in the BIM-based safety monitoring system can be referred to the corresponding process in the above method embodiments.

[0059] For ease of understanding, an example is provided below: A BIM-based safety monitoring system for power electromechanical equipment includes: The digital twin construction module is used to construct a digital twin that maps to the physical power system by integrating the three-dimensional geometric model, thermophysical property model, real-time electrical topology model, and control logic model of the power equipment. The data synchronization module is used to acquire multi-source heterogeneous data from the physical power system and synchronize the multi-source heterogeneous data to the digital twin. The multi-source heterogeneous data includes the collected equipment casing temperature, load current, environmental parameters, as well as the acquired protection signals and switch status. The predictive maintenance analysis module is used to solve the inverse problem of heat conduction in a digital twin at a preset first fixed cycle, based on the synchronized equipment casing temperature, load current, and environmental parameters, combined with a thermophysical property model. This allows for the derivation of the predicted temperature field and heat source power inside the equipment. The module then compares the predicted temperature field with a preset equipment insulation heat resistance threshold to assess the equipment's thermal risk level. When a warning or alarm level is reached, the module generates and outputs a predictive maintenance work order containing the equipment's location and predicted hotspot information. The real-time monitoring and early warning module is used to process multi-source heterogeneous data synchronized to the digital twin in real time and match it with the preset early fault model. If a match is found, the module generates and outputs the corresponding fault warning. At the same time, it monitors protection signals and switch status in real time. When an event that meets the preset electrical fault judgment conditions is detected, the self-healing process is triggered. The autonomous healing decision module is used to determine the fault range based on the real-time electrical topology model in the digital twin after the autonomous healing process is triggered, generate multiple feasible electrical isolation schemes, and determine the optimal scheme from multiple electrical isolation schemes based on preset evaluation criteria. The instruction execution module is used to perform safety logic verification on the optimal solution, generate and issue a sequence of control instructions to the physical system to perform fault isolation and power restoration to non-faulty sections.

[0060] The third embodiment of this application provides a computer device, which may include a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement a BIM-based safety monitoring method for power electromechanical equipment.

[0061] The memory can communicate with the processor via a communication bus, which can be an address bus, a data bus, a control bus, etc.

[0062] Additionally, the memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device.

[0063] Furthermore, the processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0064] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A BIM-based safety monitoring method for power electromechanical equipment, characterized in that, include: By integrating the three-dimensional geometric model, thermophysical property model, real-time electrical topology model, and control logic model of power equipment, a digital twin that maps the physical power system is constructed. Acquire multi-source heterogeneous data of the physical power system and synchronize the multi-source heterogeneous data to the digital twin; the multi-source heterogeneous data includes the collected equipment casing temperature, load current, environmental parameters, and the acquired protection signals and switch status; With a preset first fixed period, in the digital twin, based on the synchronized device casing temperature, load current and environmental parameters, and combined with the thermophysical property model, the inverse problem of heat conduction is solved to deduce the predicted temperature field and heat source power inside the device. The predicted temperature field is compared with the preset equipment insulation heat resistance threshold to assess the equipment thermal risk level. When the warning or alarm level is reached, a predictive maintenance work order containing equipment location and predicted hot spot information is generated and output. The multi-source heterogeneous data synchronized to the digital twin is processed in real time and matched with a preset early fault model. If a match is found, a corresponding fault warning is generated and output. At the same time, the protection signals and switch status are monitored in real time. When an event that meets the preset electrical fault judgment conditions is detected, the self-healing process is triggered. After the autonomous healing process is triggered, the fault range is determined based on the real-time electrical topology model in the digital twin, and multiple feasible electrical isolation schemes are generated. The optimal scheme is determined from the multiple electrical isolation schemes based on preset evaluation criteria. After safety logic verification, a sequence of control commands is generated and sent to the physical system to perform fault isolation and restore power supply to non-faulty sections.

2. The method for safety monitoring of power electromechanical equipment based on BIM according to claim 1, characterized in that, The steps for solving the inverse problem of heat conduction using the aforementioned thermophysical property model, and for retrieving the predicted temperature field and heat source power inside the equipment, specifically include: The average temperature of the housing is calculated based on the readings of multiple temperature sensors arranged on the housing. Estimate the reference heating power of the equipment based on the collected effective value of the load current; A heat conduction equation is established with the internal hot spot temperature of the equipment as the unknown core and the average temperature of the outer shell and the ambient temperature as boundary conditions. An optimization algorithm is used to iteratively adjust the assumed value of the internal heat source power to minimize the error between the shell temperature distribution calculated based on the assumed value and the measured temperature distribution using the heat conduction equation. The corresponding heat source power and the calculated internal temperature at this point are the inversion results.

3. The method for safety monitoring of power electromechanical equipment based on BIM according to claim 2, characterized in that, The specific steps for assessing the thermal risk level of equipment and generating predictive maintenance work orders include: The temperature rise margin is obtained by calculating the difference between the internal hot spot temperature and the maximum allowable temperature of the equipment insulation material. Calculate the rate of change of the internal hotspot temperature between the current cycle and the previous cycle to obtain the temperature rise rate; When the temperature rise margin is lower than the first threshold or the temperature rise rate is positive and exceeds the set rate, it is determined to be a warning level; When the temperature rise margin is lower than a lower second threshold or the internal hot spot temperature reaches the allowable limit, it is determined to be an alarm level; In the three-dimensional geometric model of the digital twin, the corresponding equipment model is highlighted with color according to the risk level, and a structured work order containing the equipment's unique identifier, predicted hot spot locations, suggested maintenance items, and recommended time windows is automatically generated.

4. The BIM-based safety monitoring method for power electromechanical equipment according to claim 1, characterized in that, The step of determining the optimal solution from multiple electrical isolation schemes based on preset evaluation criteria specifically includes: Centered on the fault point, all feasible electrical isolation schemes are simulated in parallel within the digital twin. For each electrical isolation scheme, online simulation is performed and the comprehensive loss function value is calculated. The electrical isolation scheme with the smallest comprehensive loss function value is selected as the optimal scheme.

5. A BIM-based safety monitoring method for power electromechanical equipment according to claim 4, characterized in that, The specific steps for performing online simulation and calculating the comprehensive loss function value for each of the aforementioned electrical isolation schemes include: The switching operation of the corresponding electrical isolation scheme is simulated and executed in the digital twin copy to update the electrical topology connection state; Fast power flow estimation is performed based on the updated electrical topology to obtain the reconstructed current values ​​of each branch and the voltage values ​​of each node. Based on the reconstructed current value and combined with the equipment thermal model, the new heat generation power and hot spot temperature of the key equipment can be quickly estimated, and potential thermal risk indicators can be calculated. The load loss caused by the corresponding electrical isolation scheme is statistically analyzed, and the load loss index is calculated by combining the preset load importance weights. The number of switches required to be operated and the physical distance of the operation path for the corresponding electrical isolation scheme are statistically analyzed, and the operation complexity index is calculated. Based on the load loss index, operational complexity index, and potential thermal risk index, a weighted summation algorithm is applied to calculate the comprehensive loss function value of the corresponding electrical isolation scheme.

6. A BIM-based safety monitoring method for power electromechanical equipment according to claim 1, characterized in that, The safety monitoring method also includes: After the equipment completes maintenance according to the predictive maintenance work order, the actual location and severity of the defects in the maintenance record are obtained. The actual defect information is compared with the hotspot locations and temperatures predicted by the digital twin inversion, and the prediction deviation is calculated. Based on the prediction deviation, the corresponding parameters of the thermophysical property model in the digital twin are adjusted in reverse to narrow the gap between the virtual model and the physical entity.

7. A BIM-based safety monitoring method for power electromechanical equipment according to claim 5, characterized in that, The safety monitoring method also includes: Record the actual handling results of each failure event, including the actual isolation range, load recovery time, and whether it caused subsequent problems; The actual handling results are compared and analyzed with the comprehensive loss function values ​​of each candidate electrical isolation scheme during online simulation in the digital twin; When the actual optimal solution differs from the solution selected in the simulation, or when there is a systematic deviation between the simulation evaluation index and the actual handling result, the weight coefficients of load loss, operational complexity, and potential thermal risk in the comprehensive loss function are iteratively adjusted.

8. A BIM-based safety monitoring method for power electromechanical equipment according to claim 5, characterized in that, The safety monitoring method also includes: After generating the predictive maintenance work order or the fault warning, the thermal risk level or fault warning level information of the corresponding equipment is synchronized to the decision logic of the autonomous healing process. When calculating the load loss index of the comprehensive loss function value, the corresponding importance weight coefficient is temporarily increased for the loads carried by equipment at high-risk or early warning levels. When evaluating electrical isolation schemes, priority should be given to ensuring power supply continuity for loads with high weighting factors.

9. A BIM-based safety monitoring system for power electromechanical equipment, characterized in that, Performing the BIM-based safety monitoring method for power electromechanical equipment as described in any one of claims 1 to 7 includes: The digital twin construction module is used to construct a digital twin that maps to the physical power system by integrating the three-dimensional geometric model, thermophysical property model, real-time electrical topology model, and control logic model of the power equipment. The data synchronization module is used to acquire multi-source heterogeneous data of the physical power system and synchronize the multi-source heterogeneous data to the digital twin; the multi-source heterogeneous data includes the collected equipment casing temperature, load current, environmental parameters, and the acquired protection signals and switch status; The predictive maintenance analysis module is used to solve the inverse problem of heat conduction in the digital twin at a preset first fixed period, based on the synchronized equipment casing temperature, load current, and environmental parameters, combined with the thermophysical property model, to deduce the predicted temperature field and heat source power inside the equipment; and to evaluate the thermal risk level of the equipment by comparing the predicted temperature field with the preset equipment insulation heat resistance threshold. When the warning or alarm level is reached, a predictive maintenance work order containing equipment location and predicted hotspot information is generated and output. The real-time monitoring and early warning module is used to process the multi-source heterogeneous data synchronized to the digital twin in real time and match it with the preset early fault model. If a match is found, a corresponding fault warning is generated and output. At the same time, the module monitors the protection signals and switch status in real time. When an event that meets the preset electrical fault judgment conditions is detected, the self-healing process is triggered. The autonomous healing decision module is used to determine the fault range based on the real-time electrical topology model in the digital twin after the autonomous healing process is triggered, generate multiple feasible electrical isolation schemes, and determine the optimal scheme from the multiple electrical isolation schemes based on preset evaluation criteria. The instruction execution module is used to perform safety logic verification on the optimal solution, generate and send a sequence of control instructions to the physical system to perform fault isolation and power restoration to non-faulty sections.

10. A computer device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the BIM-based safety monitoring method for power electromechanical equipment as described in any one of claims 1 to 7.

Images