Comprehensive safety inspection method for smart power plant

Abstract

The invention relates to the technical field of power plant inspection, and discloses a comprehensive safety inspection method for a smart power plant, which comprises the following steps: S1, arranging various sensors and pulse signal devices on each power device in the power plant, regularly collecting operation data and pulse reflection signal frequency spectrum of each power device, and establishing a normal operation state set of the devices; and S2, generating a visual three-dimensional map of the power plant according to the data of the sensor and the pulse reflection signal spectrum, marking an image model and a number of each device in the map, and mapping the accurate position of the device model in an actual scene to the map in an equal-proportion splicing mode. By introducing the equipment health index and residual life prediction model, the running state of the equipment is evaluated in real time, potential faults are predicted in combination with historical data, the service life of the equipment is prolonged, fault prediction is more accurate, predictive maintenance tasks are generated in advance, and serious faults are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of power plant inspection, and in particular to a comprehensive safety inspection method for a smart power plant. Background Art

[0002] With the growth of modern industry and energy demand, the scale and complexity of power plants are increasing. The reliability and safety of equipment have become crucial. In order to ensure the safe operation of power plants, the inspection of power equipment is an indispensable part of power plant operation and maintenance. Traditional power plant inspection methods mainly rely on manual inspections or drone inspections. However, due to the large number of power plant equipment, wide distribution range, and complex inspection environment, this method has many limitations.

[0003] The existing inspection method mentioned in the patent document of patent number CN117236916B can collect equipment operation data and generate alarms through abnormal detection, but it mainly relies on the monitoring of the real-time status of the equipment and lacks a prediction mechanism for future equipment failures. The existing technology cannot effectively combine the historical data of the equipment with the current operating status to evaluate the health status and remaining life of the equipment, so it is difficult to generate predictive maintenance tasks in advance, resulting in sudden failure of the equipment under unforeseen circumstances, causing unexpected downtime and equipment damage;

[0004] The traditional power plant inspection system collects equipment operation data and uploads it to the cloud for processing. This method improves the monitoring breadth of the equipment, but at the same time brings the burden of data transmission. The comparative document CN117236916B mentions the collection and analysis of equipment data, but does not optimize data screening and transmission, which can easily lead to excessive processing pressure on the cloud. Especially when the amount of equipment data is large or there are many inspection devices, the system will be subject to unnecessary load, affecting real-time performance and data processing efficiency;

[0005] In the prior art, the inspection path is usually a pre-set fixed task. When multiple devices are abnormal at the same time, the existing inspection system cannot intelligently adjust the task priority and inspection path, resulting in inefficient fault handling. The comparative document CN117236916B uses a fixed inspection task generation and allocation mechanism. When multiple devices are abnormal, it is difficult for the system to dynamically adjust the task priority and path according to the importance of the equipment and the severity of the fault, resulting in a prolonged fault handling response time.

[0006] Therefore, those skilled in the art provide a comprehensive safety inspection method for a smart power plant to solve the problems raised in the above background technology. Summary of the invention

[0007] In view of the deficiencies in the prior art, the present invention provides a comprehensive safety inspection method for a smart power plant to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: A comprehensive safety inspection method for a smart power plant, comprising the following steps: S1, deploying a variety of sensors and pulse signal devices on various power equipment in the power plant, regularly collecting operating data and pulse reflection signal spectrum of each power equipment, and establishing a normal operating status set of the equipment;

[0009] S2. Generate a visualized three-dimensional map of the power plant based on the data from the sensor and the pulse reflection signal spectrum, and mark the image model and number of each device on the map, and map the precise position of the device model in the actual scene to the map by proportional splicing;

[0010] S3. Inspection path points are set based on the three-dimensional map, and fixed inspection tasks are generated based on the priority and importance of the equipment operation status. The fixed inspection tasks calculate the estimated inspection time based on the distance between adjacent inspection path points and the standard inspection speed, and generate the inspection cycle;

[0011] S4. In the process of executing fixed inspection tasks, the health status of the equipment is dynamically analyzed based on the real-time data of the equipment and the AI ​​algorithm. If the equipment status is abnormal, the system will automatically generate random inspection tasks and adjust the priority and inspection path of the inspection equipment;

[0012] S5. Use edge computing technology to preliminarily process data in inspection equipment, filter out key data through intelligent algorithms, and upload data to the cloud platform regularly;

[0013] S6. Introduce an adaptive compression algorithm to compress the collected data, and adjust the compression ratio according to the importance and real-time nature of the data;

[0014] S7. After the inspection task is completed, the equipment status image model is generated through the data of the sensor and the pulse reflection signal, and it is compared with the equipment image model in the three-dimensional visualized map of the power plant. By comparing the degree of deformation of the equipment, its operating status is judged. If the deformation exceeds the preset deviation threshold, an equipment abnormality alarm is generated;

[0015] S8. Combine the real-time status data of the equipment with the historical inspection data, use big data analysis and AI models to predict potential equipment failures, generate predictive maintenance tasks, and perform preventive repairs before failures. If an abnormality occurs in the equipment, the system will automatically dispatch spare inspection equipment to take over the inspection task.

[0016] S9. If an equipment abnormality alarm is generated, the system automatically sends maintenance information to the maintenance team and matches the supporting power equipment with the same function according to the number of the equipment image model;

[0017] S10. In the process of data collection, transmission and storage, end-to-end encryption technology is adopted, and distributed storage and permission control mechanisms are introduced to limit access to sensitive data, so as to record data operations and access logs through blockchain technology.

[0018] Preferably, in the step S3, the calculation formula for the estimated inspection time is:

[0019]

[0020] Among them, t is the expected inspection time, d i，j is the distance between path points i and j, and V is the inspection speed.

[0021] Preferably, in the step S4, the priority of the inspection task is calculated by the following formula:

[0022]

[0023] Among them, P is the task priority, F(t d ,s d ) is the fault severity and time delay function, T c is the remaining time of the inspection cycle, W c is the weight of the task.

[0024] Preferably, in step S6, the compression algorithm formula is:

[0025] C=f(D high , D low ,α),

[0026] Among them, C is the compression ratio, D high For high priority data, D low is low priority data, and α is the compression parameter.

[0027] Preferably, in step S7, the calculation formula of the deviation value is:

[0028] H=∑|xX|,

[0029] Among them, H is the deformation deviation value, x is the real-time status of the equipment, and X is the standard status of the equipment.

[0030] Preferably, in the step S2, the pulse reflection signal data of each power device is combined with the actual position of the device in the power plant to generate a three-dimensional map model, and the process includes pulse reflection signal spectrum analysis, generation and splicing of device image models and three-dimensional space mapping;

[0031] The pulse reflection signal spectrum analysis: The pulse reflection signal is used to obtain the physical characteristics of the device. First, the peak and trough positions in the pulse reflection signal spectrum are extracted, and the device image model is generated based on the information. The algorithm formula is: signal strength difference = S peak -S trough , where S peak Represents the peak in the pulse reflection signal, S trough Indicates the trough;

[0032] The generation of the three-dimensional image model of the device: mapping the pulse signal reflection characteristics of the device to a three-dimensional rectangular coordinate system, generating a three-dimensional model of the device, setting the position vector (x i ,y i , z i ), whose device profile is calculated by the following formula:

[0033] (x i ,y i , z i )=f(S peak , S trough , d), where d is the propagation distance of the pulse signal and f is the mapping function between the shape of the device and the distance;

[0034] Equipment splicing of the three-dimensional map: splicing the three-dimensional image models of the power equipment according to their actual physical positions in the power plant to construct a complete three-dimensional map. The three-dimensional models of each device are positioned according to their relative positions. The position offset calculation formula is:

[0035] (x map ,y map , z map )=(x i +Δx,y i +Δy,z i +Δz),

[0036] Among them, (x i ,y i , z i ) is the initial coordinate of the equipment image model, Δx, Δy, Δz are the position offsets of the equipment relative to the power plant;

[0037] The device scaling: Each device needs to be scaled proportionally. The scaling factor λ of the device is determined by the actual physical size of the device and its signal reflection characteristics: Among them, L real is the actual physical size of the device, L signal is the device size derived from signal reflection analysis;

[0038] The three-dimensional coordinate mapping: The three-dimensional map of the power plant is a collection of equipment image models. Through the above steps, the three-dimensional coordinates (x map ,y map , z map ), and finally combine the equipment into a complete 3D visualization map of the power plant. The final construction formula of the 3D map is:

[0039]

[0040] Where n is the total number of equipment in the power plant, λ i The scaling factor for each device.

[0041] Preferably, in the step S5, the data collected by the device is processed and screened in real time through the edge computing node, and the algorithm formula includes an anomaly detection algorithm, a priority data screening algorithm, an intelligent data compression algorithm, an edge node processing time optimization formula, and an edge computing and cloud collaborative optimization formula;

[0042] In the anomaly detection algorithm, the edge node monitors the sensor data in real time and determines whether the data is within the normal range by comparing the real-time data D real Compared with the preset normal operating state range [D min , D max ] to determine whether the data is abnormal, the formula is: D real ∈[D min , D max ],

[0043] If D real Out of range, that is, the following conditions are met:

[0044] D real <D min or D real >D max ,

[0045] The priority data screening algorithm, if the edge node needs to decide whether data needs to be uploaded to the cloud, screens based on the priority of the data. The priority P of each piece of data can be comprehensively evaluated according to the abnormality of the data and the importance of the device. The formula is:

[0046] Among them, w1 is the weight of the data abnormality, D avg is the average data value during normal operation, I device is the importance index of the equipment;

[0047] In the intelligent data compression algorithm, the edge computing node will compress the data to be transmitted, and the compression ratio C is dynamically adjusted according to the importance P of the data and the historical fluctuation V. The formula is: Among them, α is the compression adjustment parameter, P compress is the compression threshold.

[0048] Preferably, the edge node processing time optimization formula completes the screening within a reasonable time and defines the processing time T of each data p , the relationship between data size S and computing resources R, the formula is: If the processing time T p Exceeding a certain threshold T MAX , then reduce the data processing accuracy or increase the allocation of computing resources;

[0049] The edge computing and cloud collaborative optimization formula, the data transmission volume between the edge node and the cloud platform D transmit Based on the importance of data P and the storage capacity of edge nodes S edge Dynamic adjustment, the formula is: Among them, D edge Indicates the amount of data stored in the edge node, S edge Indicates the available storage space of the edge node.

[0050] Preferably, in the step S8, rapid response measures are taken when an abnormality occurs in the equipment, including equipment health assessment and fault prediction model and remaining life prediction model;

[0051] The equipment health assessment and fault prediction model, equipment health assessment is based on the combination of historical operation data and real-time data, and the health index H(t) of the equipment is estimated through the prediction model. This index is used to represent the equipment status and the possibility of fault. Its algorithm formula is:

[0052]

[0053] Among them, D real (t) is the actual data value of the device at time t, D avg is the average data value when the equipment is operating normally, V(t) is the fluctuation amplitude of the equipment data at time t, and F past is the number of equipment failures, F total is the total number of equipment failures, w1, w2, w3 are weight coefficients;

[0054] If the health index H(t) is less than the preset threshold H threshold , that is: H(t) <H threshold ,

[0055] The system triggers a predictive maintenance task, reminding maintenance personnel to inspect and preventively repair the equipment.

[0056] Preferably, the remaining service life prediction model calculates the remaining service life of the equipment after the equipment health assessment. The remaining service life is an estimated value of the time that the equipment can operate normally in the future without maintenance. The calculation of the remaining service life is based on the degradation model, and the formula is as follows:

[0057] Where H(t) is the health index of the equipment, is the degradation rate of the equipment health index.

[0058] The present invention provides a comprehensive safety inspection method for a smart power plant, which has the following beneficial effects:

[0059] 1. The present invention introduces the equipment health index and the remaining life prediction model to evaluate the operating status of the equipment in real time, predict potential failures in combination with historical data, increase the service life of the equipment, make failure prediction more accurate, generate predictive maintenance tasks in advance, and avoid the occurrence of serious failures.

[0060] 2. The present invention processes data locally in the inspection equipment through edge computing, transmits key data, and reduces unnecessary data transmission and cloud load. At the same time, the system introduces intelligent data screening and compression algorithms to dynamically adjust data transmission and storage priorities to improve the efficiency of overall data management.

[0061] 3. The present invention uses AI algorithms and task priority calculation formulas to intelligently adjust the inspection path and priority in real time according to the importance of the equipment and the severity of the fault when multiple devices are abnormal, ensuring the system's rapid response and flexible scheduling when dealing with emergency faults, and avoiding fault handling delays caused by fixed path inspections. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0063] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0064] The present invention is described in detail below in conjunction with the accompanying drawings:

[0065] Example:

[0066] Please see attached Figure 1, an embodiment of the present invention provides a comprehensive safety inspection method for a smart power plant, comprising the following steps: S1, deploying a variety of sensors and pulse signal devices on various power equipment in the power plant, regularly collecting operating data and pulse reflection signal spectrum of each power equipment, and establishing a normal operating status set of the equipment;

[0067] S2. Generate a visualized three-dimensional map of the power plant based on the data from the sensor and the pulse reflection signal spectrum, and mark the image model and number of each device on the map, and map the precise position of the device model in the actual scene to the map by proportional splicing;

[0068] S3. Inspection path points are set based on the three-dimensional map, and fixed inspection tasks are generated based on the priority and importance of the equipment operation status. The fixed inspection tasks calculate the estimated inspection time based on the distance between adjacent inspection path points and the standard inspection speed, and generate the inspection cycle;

[0069] S4. In the process of executing fixed inspection tasks, the health status of the equipment is dynamically analyzed based on the real-time data of the equipment and the AI ​​algorithm. If the equipment status is abnormal, the system will automatically generate random inspection tasks and adjust the priority and inspection path of the inspection equipment;

[0070] S5. Use edge computing technology to preliminarily process data in inspection equipment, filter out key data through intelligent algorithms, and upload data to the cloud platform regularly;

[0071] S6. Introduce an adaptive compression algorithm to compress the collected data, and adjust the compression ratio according to the importance and real-time nature of the data;

[0072] S7. After the inspection task is completed, the equipment status image model is generated through the data of the sensor and the pulse reflection signal, and it is compared with the equipment image model in the three-dimensional visualized map of the power plant. By comparing the degree of deformation of the equipment, its operating status is judged. If the deformation exceeds the preset deviation threshold, an equipment abnormality alarm is generated;

[0073] S8. Combine the real-time status data of the equipment with the historical inspection data, use big data analysis and AI models to predict potential equipment failures, generate predictive maintenance tasks, and perform preventive repairs before failures. If an abnormality occurs in the equipment, the system will automatically dispatch spare inspection equipment to take over the inspection task.

[0074] S9. If an equipment abnormality alarm is generated, the system automatically sends maintenance information to the maintenance team and matches the supporting power equipment with the same function according to the number of the equipment image model;

[0075] S10. In the process of data collection, transmission and storage, end-to-end encryption technology is adopted, and distributed storage and permission control mechanisms are introduced to limit access to sensitive data, so as to record data operations and access logs through blockchain technology.

[0076] The benefit of S1 is that it provides a data basis for subsequent equipment status analysis and three-dimensional map construction, which helps to detect potential problems in a timely manner and improve the comprehensiveness and accuracy of equipment monitoring; the benefit of S2 is that it displays the equipment distribution of the power plant through proportional splicing, which is convenient for management and maintenance, can better identify the spatial correlation of equipment, and improve the comprehensiveness and operability of inspections; the benefit of S3 is that it calculates the estimated inspection time and inspection cycle based on the distance and inspection speed, ensures the efficiency and timeliness of the inspection work, and ensures the real-time monitoring and stable operation of the equipment; the benefit of S4 is to improve the flexibility and response speed of the inspection system, ensure that when a fault occurs, the system can make rapid adjustments to avoid response delays caused by fixed-path inspections; the benefit of S5 is to reduce data transmission costs and enhance real-time performance , ensuring that inspection data can be processed faster; the benefit of S6 is to optimize resource utilization and further improve the efficiency of data transmission and storage; the benefit of S7 is to judge the equipment operation status through deformation deviation detection, to ensure that problems can be detected in time when the equipment is abnormal, to generate early warnings, to prevent the expansion of equipment failures, and to improve the safety of power plant equipment; the benefit of S8 is to extend the service life of equipment, reduce the risk of sudden downtime, and improve the safety and efficiency of the overall operation and maintenance of power plants; the benefit of S9 is to ensure the stability of power equipment operation and the continuity of power plant production by matching the takeover function of supporting power equipment with the same function, and to avoid losses caused by equipment downtime; the benefit of S10 is to effectively limit access rights to sensitive data, improve data security, and ensure the compliance and security of power plant data management.

[0077] In step S3, the calculation formula for the estimated inspection time is: Among them, t is the expected inspection time, d i，j is the distance between path points i and j, and V is the inspection speed;

[0078] In step S4, the priority of the inspection task is calculated using the following formula: Among them, P is the task priority, F(t d ,s d ) is the fault severity and time delay function, T c is the remaining time of the inspection cycle, W c is the weight of the task;

[0079] In step S6, the compression algorithm formula is: C = f(D high , D low , α), where C is the compression ratio, D high For high priority data, Dlow is low priority data, α is the compression parameter;

[0080] In step S7, the calculation formula of the deviation value is: H=∑|xX|, where H is the deformation deviation value, x is the real-time state of the equipment, and X is the standard state of the equipment.

[0081] The calculation formula of the estimated inspection time is used to calculate the time required for the inspection equipment to move between the various path points in the power plant. With the known inspection speed and the distance between the path points, the time required for the inspection can be reasonably estimated, and the inspection plan can be formulated to ensure that the inspection task is completed within the specified time period, thereby improving the inspection efficiency;

[0082] The calculation formula of inspection task priority is used to calculate the priority of inspection tasks, especially when multiple devices fail at the same time. Resources can be reasonably scheduled through priority sorting. The priority of tasks is determined by the severity and delay time of the failure. Combined with the remaining time of the inspection cycle and the task weight, the system can intelligently adjust the inspection path and task sequence to ensure that key equipment and serious failures are handled first, thereby improving the flexibility and response speed of failure response.

[0083] The compression algorithm formula dynamically adjusts the data compression ratio according to the data priority and compression parameters to ensure the accuracy of data transmission, and compresses low-priority data at a higher ratio to reduce the transmission burden. Through intelligent compression, the system can effectively reduce data transmission and storage pressure, optimize bandwidth usage, and ensure the accuracy and transmission quality of key data.

[0084] The calculation formula of the deformation deviation value is used to calculate the deviation value between the real-time state of the equipment and the standard state. By comparing the actual state of the equipment with the standard state, it is evaluated whether the equipment has deformation or abnormality. If the deviation value exceeds the preset threshold, the system will generate an equipment abnormality alarm.

[0085] In step S2, the pulse reflection signal data of each power device is combined with the actual position of the device in the power plant to generate a three-dimensional map model, and the process includes pulse reflection signal spectrum analysis, device image model generation and splicing, and three-dimensional space mapping;

[0086] Pulse reflection signal spectrum analysis: Pulse reflection signals are used to obtain the physical characteristics of the device. First, the peak and trough positions in the pulse reflection signal spectrum are extracted, and the device image model is generated based on the information. The algorithm formula is: Signal strength difference = S peak -S trough , where S peak Represents the peak in the pulse reflection signal, S trough Indicates the trough;

[0087] Generation of a three-dimensional image model of the equipment: Map the pulse signal reflection characteristics of the equipment into a three-dimensional rectangular coordinate system to generate a three-dimensional model of the equipment. The position vector (x i ,y i , z i ), whose device profile is calculated by the following formula:

[0088] (x i ,y i , z i )=f(S peak , S trough , d), where d is the propagation distance of the pulse signal and f is the mapping function between the shape of the device and the distance;

[0089] Equipment stitching of 3D maps: stitch the 3D image models of power equipment according to their actual physical locations in the power plant to build a complete 3D map. The 3D models of each device are positioned according to their relative positions. The position offset calculation formula is:

[0090] (x map ,y map , z map )=(x i +Δx,y i +Δy,z i +Δz),

[0091] Among them, (x i ,y i , z i ) is the initial coordinate of the equipment image model, Δx, Δy, Δz are the position offsets of the equipment relative to the power plant;

[0092] Scaling between devices: Each device needs to be scaled proportionally. The scaling factor λ of the device is determined by the actual physical size of the device and its signal reflection characteristics: Among them, L real is the actual physical size of the device, L signal is the device size derived from signal reflection analysis;

[0093] Three-dimensional coordinate mapping: The three-dimensional map of the power plant is a collection of equipment image models. Through the above steps, the three-dimensional coordinates (x map ,y map , z map ), and finally combine the equipment into a complete 3D visualization map of the power plant. The final construction formula of the 3D map is:

[0094]

[0095] Where n is the total number of equipment in the power plant, λ iThe scaling factor for each device.

[0096] The pulse reflection signal spectrum analysis formula is used to analyze the peaks and troughs in the pulse reflection signal spectrum to extract the physical characteristics of the device. The characteristics reflect the geometric shape and surface characteristics of the device. Through the difference between the peaks and troughs, the system identifies the size and deformation of the device and generates an image model of the device based on this. The signal strength difference provides basic information for the image modeling of the device, making the subsequent three-dimensional model more accurate.

[0097] The function of the generation formula of the equipment three-dimensional image model is to map the pulse reflection signal characteristics of the equipment into a three-dimensional coordinate system to generate a three-dimensional image model of the equipment. By combining the peaks and troughs of the pulse signal with the signal propagation distance, the position and shape characteristics of the equipment in space are determined. The process converts the physical characteristics of the equipment into three-dimensional space coordinates, so that each device can be presented in the power plant map in the form of a three-dimensional graphic;

[0098] The function of the formula for equipment splicing in the 3D map is to match the generated 3D equipment model with the actual physical location in the power plant. By calculating the coordinate offset of the equipment, each equipment model is accurately positioned in the space of the power plant to ensure that the location of the equipment is consistent with the layout in the actual power plant, so that the 3D models of multiple equipment can be accurately spliced ​​together to build a complete 3D map of the power plant.

[0099] The function of the scaling formula between devices is to scale the device model in equal proportion. There is a deviation between the physical size of the device and the size obtained through pulse signal analysis. The scaling factor can ensure that the device is displayed in the three-dimensional map with the real size ratio. The scaling process ensures the accuracy of the relative size and spatial layout between devices in the map, making the three-dimensional map more in line with the actual scene;

[0100] The three-dimensional coordinate mapping formula is used to construct a complete three-dimensional map of the power plant. The three-dimensional coordinates and scaling of the equipment are combined to generate a complete three-dimensional visualization model. Through this formula, the spatial position and size of the equipment in the power plant are accurately mapped to the three-dimensional map, so that the inspection system can monitor and analyze the equipment based on the map. Ultimately, the three-dimensional map provides an intuitive visualization tool for the inspection of the power plant, improving the efficiency of power plant equipment management.

[0101] In step S5, the data collected by the device is processed and screened in real time through the edge computing node. The algorithm formula includes anomaly detection algorithm, priority data screening algorithm, intelligent data compression algorithm, edge node processing time optimization formula and edge computing and cloud collaborative optimization formula;

[0102] Anomaly detection algorithm, the edge node monitors the sensor data in real time, determines whether the data is within the normal range, and compares the real-time data D realCompared with the preset normal operating state range [D min , D max ] to determine whether the data is abnormal, the formula is: D real ∈[D min , D max ],

[0103] If D real Out of range, that is, the following conditions are met: D real <D min or D real >D max ,

[0104] Priority data screening algorithm: If the edge node needs to decide whether data needs to be uploaded to the cloud, it will screen based on the priority of the data. The priority P of each piece of data can be comprehensively evaluated based on the degree of abnormality of the data and the importance of the device. The formula is: Among them, w1 is the weight of the data abnormality, D avg is the average data value during normal operation, I device is the importance index of the equipment;

[0105] Intelligent data compression algorithm: Edge computing nodes compress the data to be transmitted. The compression ratio C is dynamically adjusted according to the importance P of the data and the historical fluctuation V. The formula is:

[0106] Among them, α is the compression adjustment parameter, P compress is the compression threshold;

[0107] Edge node processing time optimization formula, complete the screening within a reasonable time, define the processing time T of each data p , the relationship between data size S and computing resources R, the formula is: If the processing time T p Exceeding a certain threshold T MAX , then reduce the data processing accuracy or increase the allocation of computing resources;

[0108] Edge computing and cloud collaborative optimization formula, data transmission volume D between edge nodes and cloud platforms transmit Based on the importance of data P and the storage capacity of edge nodes S edge Dynamic adjustment, the formula is:

[0109] Among them, D edge Indicates the amount of data stored in the edge node, S edge Indicates the available storage space of the edge node.

[0110] The function of the anomaly detection algorithm formula is to monitor the data collected by the equipment sensors in real time and determine whether the data exceeds the normal operating range. By setting the numerical range of normal operation of the equipment, the system quickly identifies the abnormal situation of the equipment. Any data exceeding the range will be marked as abnormal, triggering an alarm or further processing measures. Its anomaly detection algorithm ensures that equipment failures can be detected in time to prevent the potential problems of the equipment from deteriorating.

[0111] The priority data screening algorithm formula is used to prioritize data based on the degree of abnormality of the data and the importance of the equipment. The degree of abnormality reflects the deviation of the equipment status, and the importance index of the equipment determines the criticality of the equipment in the entire power plant. High-priority data will be uploaded to the cloud for further processing. This mechanism may optimize the system's management of data, ensure that high-priority data is processed in a timely manner, and avoid low-priority data occupying too many system resources;

[0112] The function of the intelligent data compression algorithm formula is to dynamically adjust the data compression ratio C according to the importance P of the data. High-importance data will be compressed less to ensure the accuracy of data transmission, and low-importance data will be proportionally compressed to reduce the transmission burden. Through intelligent compression, the system can reduce the pressure on network bandwidth and storage resources while maintaining the integrity of key data;

[0113] The edge node processing time optimization formula is used to calculate the time required for the edge node to process each piece of data. If the processing time exceeds the set threshold, the system will reduce the accuracy of data processing or increase the allocation of computing resources to ensure processing efficiency. When processing a large amount of data, this formula can optimize the allocation of system resources so that the edge computing node can complete data screening and processing within a reasonable time, avoiding data accumulation or processing delays;

[0114] The edge computing and cloud collaborative optimization formula is used to dynamically adjust the amount of data transmission between edge nodes and the cloud. The amount of data transmitted is determined based on the importance P of the data and the storage capacity of the edge nodes. If the storage space is limited, the system will give priority to transmitting data with higher importance and reduce the transmission of low-priority data.

[0115] In step S8, rapid response measures are taken when equipment abnormalities occur, including equipment health assessment and failure prediction model and remaining life prediction model;

[0116] Equipment health assessment and fault prediction model. Equipment health assessment is based on the combination of historical operation data and real-time data. The health index H(t) of the equipment is estimated through the prediction model. This index is used to represent the equipment status and the possibility of fault. Its algorithm formula is:

[0117]

[0118] Among them, D real (t) is the actual data value of the device at time t, D avg is the average data value when the equipment is operating normally, V(t) is the fluctuation amplitude of the equipment data at time t, and F past is the number of equipment failures, F total is the total number of equipment failures, w1, w2, w3 are weight coefficients;

[0119] If the health index H(t) is less than the preset threshold H threshold , that is: H(t) <H threshold ,

[0120] The system triggers a predictive maintenance task, reminding maintenance personnel to inspect and preventively repair the equipment;

[0121] The remaining service life prediction model calculates the remaining service life of the equipment after the equipment health assessment. The remaining service life is an estimate of the time the equipment can operate normally in the future without maintenance. The calculation of the remaining service life is based on the degradation model. The formula is as follows:

[0122] Where H(t) is the health index of the equipment, is the degradation rate of the equipment health index.

[0123] The equipment health assessment and fault prediction model formula is used to evaluate the health status of the equipment in real time. By combining the real-time data of the equipment, the fluctuation of historical data and the frequency of fault occurrence, the equipment health index is calculated. The equipment health index represents the operating status of the equipment and the potential possibility of faults. This formula can quantify the health status of the equipment and help predict whether the equipment is close to failure.

[0124] The remaining life prediction model formula is used to predict the remaining service life of the equipment, that is, the time the equipment can operate normally without maintenance. The remaining life is based on the health index of the equipment and its degradation rate. By calculating the health status of the equipment and the rate of decline of the health status, the length of time the equipment can operate normally in the future is estimated. The formula can help the maintenance team formulate a more reasonable equipment maintenance plan, ensure preventive maintenance before the equipment is about to fail, extend the service life of the equipment, and reduce the risk of unexpected downtime.

[0125] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A comprehensive safety inspection method for a smart power plant, characterized in that: The steps include: S1. Various sensors and pulse signal devices are deployed on various power equipment in the power plant to regularly collect the operating data and pulse reflection signal spectrum of each power equipment and establish a normal operating status set of the equipment; S2. Generate a visualized three-dimensional map of the power plant based on the data from the sensor and the pulse reflection signal spectrum, and mark the image model and number of each device on the map, and map the precise position of the device model in the actual scene to the map by proportional splicing; S3. Inspection path points are set based on the three-dimensional map, and fixed inspection tasks are generated based on the priority and importance of the equipment operation status. The fixed inspection tasks calculate the estimated inspection time based on the distance between adjacent inspection path points and the standard inspection speed, and generate the inspection cycle; S4. In the process of executing fixed inspection tasks, the health status of the equipment is dynamically analyzed based on the real-time data of the equipment and the AI ​​algorithm. If the equipment status is abnormal, the system will automatically generate random inspection tasks and adjust the priority and inspection path of the inspection equipment; S5. Use edge computing technology to preliminarily process data in inspection equipment, filter out key data through intelligent algorithms, and upload data to the cloud platform regularly; S6. Introduce an adaptive compression algorithm to compress the collected data, and adjust the compression ratio according to the importance and real-time nature of the data; S7. After the inspection task is completed, the equipment status image model is generated through the data of the sensor and the pulse reflection signal, and it is compared with the equipment image model in the three-dimensional visualized map of the power plant. By comparing the degree of deformation of the equipment, its operating status is judged. If the deformation exceeds the preset deviation threshold, an equipment abnormality alarm is generated; S8. Combine the real-time status data of the equipment with the historical inspection data, use big data analysis and AI models to predict potential equipment failures, generate predictive maintenance tasks, and perform preventive repairs before failures. If an abnormality occurs in the equipment, the system will automatically dispatch spare inspection equipment to take over the inspection task. S9. If an equipment abnormality alarm is generated, the system automatically sends maintenance information to the maintenance team and matches the supporting power equipment with the same function according to the number of the equipment image model; S10. In the process of data collection, transmission and storage, end-to-end encryption technology is adopted, and distributed storage and permission control mechanisms are introduced to limit access to sensitive data, so as to record data operations and access logs through blockchain technology.

2. A comprehensive safety inspection method for a smart power plant according to claim 1, characterized in that: In the step S3, the calculation formula for the estimated inspection time is: Among them, t is the expected inspection time, d i，j is the distance between path points i and j, and V is the inspection speed.

3. The comprehensive safety inspection method of a smart power plant according to claim 1 is characterized in that: In the step S4, the priority of the inspection task is calculated by the following formula: Among them, P is the task priority, F(t d ,s d ) is the fault severity and time delay function, T c is the remaining time of the inspection cycle, W c is the weight of the task.

4. The comprehensive safety inspection method of a smart power plant according to claim 1 is characterized in that: In step S6, the compression algorithm formula is: C=f(D high ,D low ,α), Among them, C is the compression ratio, D high For high priority data, D low is low priority data, and α is the compression parameter.

5. The comprehensive safety inspection method of a smart power plant according to claim 1 is characterized in that: In the step S7, the calculation formula of the deviation value is: H=∑|xX|, Among them, H is the deformation deviation value, x is the real-time status of the equipment, and X is the standard status of the equipment.

6. The comprehensive safety inspection method of a smart power plant according to claim 1 is characterized in that: In the step S2, the pulse reflection signal data of each power device is combined with the actual position of the device in the power plant to generate a three-dimensional map model, and the process includes pulse reflection signal spectrum analysis, device image model generation and splicing, and three-dimensional space mapping; The pulse reflection signal spectrum analysis: The pulse reflection signal is used to obtain the physical characteristics of the device. First, the peak and trough positions in the pulse reflection signal spectrum are extracted, and the device image model is generated based on the information. The algorithm formula is: signal strength difference = S peak -S trough , where S peak Represents the peak in the pulse reflection signal, S trough Indicates the trough; The generation of the three-dimensional image model of the device: mapping the pulse signal reflection characteristics of the device to a three-dimensional rectangular coordinate system, generating a three-dimensional model of the device, setting the position vector (x i ,y i , z i ), whose device profile is calculated by the following formula: (x i ,y i , z i )=f(S peak , S trough , d), where d is the propagation distance of the pulse signal and f is the mapping function between the shape of the device and the distance; Equipment splicing of the three-dimensional map: splicing the three-dimensional image models of the power equipment according to their actual physical positions in the power plant to construct a complete three-dimensional map. The three-dimensional models of each device are positioned according to their relative positions. The position offset calculation formula is: (x map ,y map ,z map )=(x i +Δx,y i +Δy,z i +Δz), Among them, (x i ,y i , z i ) is the initial coordinate of the equipment image model, Δx, Δy, Δz are the position offsets of the equipment relative to the power plant; The device scaling: Each device needs to be scaled proportionally. The scaling factor λ of the device is determined by the actual physical size of the device and its signal reflection characteristics: Among them, L real is the actual physical size of the device, L signal is the device size derived from signal reflection analysis; The three-dimensional coordinate mapping: The three-dimensional map of the power plant is a collection of equipment image models. Through the above steps, the three-dimensional coordinates (x map ,y map , z map ), and finally combine the equipment into a complete 3D visualization map of the power plant. The final construction formula of the 3D map is: Where n is the total number of equipment in the power plant, λ i The scaling factor for each device.

7. The comprehensive safety inspection method of a smart power plant according to claim 1 is characterized in that: In the step S5, the data collected by the device is processed and screened in real time through the edge computing node, and the algorithm formula includes an anomaly detection algorithm, a priority data screening algorithm, an intelligent data compression algorithm, an edge node processing time optimization formula, and an edge computing and cloud collaborative optimization formula; In the anomaly detection algorithm, the edge node monitors the sensor data in real time and determines whether the data is within the normal range by comparing the real-time data D real Compared with the preset normal operating state range [D min , D max ] to determine whether the data is abnormal, the formula is: D real ∈[D min , D max ], If D real Out of range, that is, the following conditions are met: D real <D min or D real >D max , The priority data screening algorithm, if the edge node needs to decide whether data needs to be uploaded to the cloud, screens based on the priority of the data. The priority P of each piece of data can be comprehensively evaluated according to the abnormality of the data and the importance of the device. The formula is: Among them, w1 is the weight of the data abnormality, D avg is the average data value during normal operation, I device is the importance index of the equipment; In the intelligent data compression algorithm, the edge computing node will compress the data to be transmitted, and the compression ratio C is dynamically adjusted according to the importance P of the data and the historical fluctuation V. The formula is: Among them, α is the compression adjustment parameter, P compress is the compression threshold.

8. A comprehensive safety inspection method for a smart power plant according to claim 7, characterized in that: The edge node processing time optimization formula completes the screening within a reasonable time and defines the processing time T of each data p , the relationship between data size S and computing resources R, the formula is: If the processing time T p Exceeding a certain threshold T MAX , then reduce the data processing accuracy or increase the allocation of computing resources; The edge computing and cloud collaborative optimization formula, the data transmission volume between the edge node and the cloud platform D transmit Based on the importance of data P and the storage capacity of edge nodes S edge Dynamic adjustment, the formula is: Among them, D edge Indicates the amount of data stored in the edge node, S edge Indicates the available storage space of the edge node.

9. The comprehensive safety inspection method of a smart power plant according to claim 1, characterized in that: In the step S8, rapid response measures are taken when an abnormality occurs in the equipment, including equipment health assessment and fault prediction model and remaining life prediction model; The equipment health assessment and fault prediction model, equipment health assessment is based on the combination of historical operation data and real-time data, and the health index H(t) of the equipment is estimated through the prediction model. This index is used to represent the equipment status and the possibility of fault. Its algorithm formula is: Among them, D real (t) is the actual data value of the device at time t, D avg is the average data value when the equipment is operating normally, V(t) is the fluctuation amplitude of the equipment data at time t, and F past is the number of equipment failures, F total is the total number of equipment failures, w1, w2, w3 are weight coefficients; If the health index H(t) is less than the preset threshold H threshold , that is: H(t) <H threshold , The system triggers a predictive maintenance task, reminding maintenance personnel to inspect and preventively repair the equipment.

10. A comprehensive safety inspection method for a smart power plant according to claim 9, characterized in that: The remaining service life prediction model calculates the remaining service life of the equipment after the equipment health assessment. The remaining service life is an estimated value of the time that the equipment can operate normally in the future without maintenance. The calculation of the remaining service life is based on the degradation model, and the formula is as follows: Among them, H(t) is the health index of the equipment, is the degradation rate of the equipment health index.

Images