Power transmission line real-time monitoring method and platform

Abstract

The invention relates to the technical field of power transmission line monitoring, and discloses a power transmission line real-time monitoring method and platform. The method comprises the following steps: establishing a multi-omics data fusion subsystem to collect metagenome, metabolome, host genome and clinical phenotype group data of a target patient; inputting the data into a flora-metabolite combined network analysis model to construct an interaction network and extracting features; generating an incidence matrix based on the features and the host genome data and calculating indexes; generating indexes through a dynamic response algorithm in combination with the clinical phenotypic data and the indexes; and outputting a curative effect prediction result by using a transfer learning framework combined with modeling. The system comprises a data acquisition module, a network analysis module, a correlation calculation module, a dynamic response module and a joint modeling module. According to the method, multiple omics data are integrated, the flora and host relation is accurately mined, intelligent prediction of the micro-ecological transplantation curative effect is achieved, powerful support is provided for IBS personalized treatment, and meanwhile data processing and safety guarantee measures are taken.

Description

Technical Field

[0001] The present invention relates to the technical field of transmission line monitoring, and in particular to a real-time monitoring method and platform for transmission lines. Background Art

[0002] In modern power systems, transmission lines, as key infrastructure for power transmission, are crucial for their safe and stable operation. With rapid economic development and the continued growth of society's demand for electricity, the scale and complexity of transmission lines are constantly increasing, and the operational risks they face are also becoming increasingly diverse.

[0003] Traditional methods for monitoring transmission lines have numerous limitations. Early inspections primarily relied on manual inspections, a method not only inefficient but also significantly constrained by factors such as geographic environment and weather conditions. For example, in complex terrain such as mountainous areas and jungles, manual inspections are difficult and costly, and real-time monitoring is difficult. Consequently, many potential faults go undetected, leading to the accumulation of potential faults and potentially serious power outages. Even with the aid of simple tools like telescopes and infrared thermometers, comprehensive and accurate information on the line's operating status cannot be obtained.

[0004] With technological advancements, some monitoring methods based on fixed sensors have begun to be used. However, these methods mostly monitor only a single parameter and fail to meet the demand for comprehensive, multi-parameter analysis of transmission lines. For example, monitoring only current or temperature makes it difficult to accurately assess the overall health of a line. When complex faults occur on a line, changes in a single parameter may not be obvious, making it easy to miss faults. Furthermore, traditional monitoring systems have weak data analysis capabilities, often simply recording and displaying monitoring data without deep data mining and analysis, making it impossible to predict the occurrence and development of faults in advance.

[0005] Furthermore, existing monitoring technologies struggle to cope with complex line topologies and variable environmental interference. Transmission lines are widely distributed, traversing diverse geographical regions and subject to varying degrees of electromagnetic interference and weather changes. For example, in environments with strong electromagnetic interference, monitoring signals are easily distorted, resulting in inaccurate monitoring data. In severe weather conditions such as heavy rain, strong winds, and lightning strikes, the operating status of the line can change dramatically, making it difficult for traditional monitoring methods to quickly and accurately assess the impact of these changes on the safe operation of the line.

[0006] In terms of data management, traditional monitoring systems also lack the proper storage and classification methods. Data is often stored in a single database without proper classification and hierarchical management, resulting in inefficient data query and retrieval. Faced with a flood of monitoring data, operations and maintenance personnel struggle to quickly access critical information and make informed decisions, impacting the efficiency and safety of transmission line operations. In summary, existing transmission line monitoring technology urgently needs improvement to meet the requirements of modern power systems for safe and reliable transmission line operation. Summary of the Invention

[0007] The purpose of the present invention is to provide a real-time monitoring method and platform for power transmission lines to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: a real-time monitoring method for a power transmission line, the method comprising:

[0009] Real-time acquisition of transmission line operating parameters using a multimodal sensor array, wherein the operating parameters include current fluctuation value, temperature gradient distribution, mechanical vibration intensity, and insulation aging index;

[0010] Inputting the operating parameters into a preset time-domain feature extraction model to identify and mark abnormal signal segments, wherein the time-domain feature extraction model dynamically adjusts the analysis window length based on the waveform characteristics of different defect modes in historical fault data;

[0011] Inputting the marked operating parameters into a preset fault evolution model to generate risk diffusion prediction results, wherein the fault evolution model performs multi-scale coupling analysis based on the line topology and environmental interference factors;

[0012] The prediction results and current operating parameters are hierarchically aggregated using a preset dynamic optimization algorithm to generate and store a real-time monitoring data set.

[0013] Preferably, the step of constructing the time domain feature extraction model includes: obtaining a historical fault data set, wherein each data in the historical fault data set is marked with a defect type and a hazard level; dividing training subsets based on the defect type and hazard level, each training subset corresponding to a fault scenario; using the training subsets to train the initial feature extraction model in parallel, until the waveform recognition accuracy of the initial feature extraction model for each fault scenario is greater than or equal to a preset first threshold, and then stopping the training to obtain an intermediate feature extraction model; inputting the historical fault data set into the intermediate feature extraction model to verify whether the feature matching degree output by the intermediate feature extraction model meets the preset error range; if so, determining the intermediate feature extraction model as the time domain feature extraction model.

[0014] Preferably, the real-time acquisition of the operating parameters of the power transmission line by the multimodal sensor array includes:

[0015] Establishing a communication link with a distributed monitoring node, wherein the distributed monitoring node is deployed at a preset key monitoring location of the transmission line;

[0016] Continuously acquiring the real-time signal of the distributed monitoring node according to a preset sampling frequency, and marking an acquisition time stamp based on the spectrum characteristics of the real-time signal;

[0017] According to the physical connection relationship of the transmission lines, the real-time signals of different monitoring locations at the same timestamp are spatially aligned to form a set of associated operating parameters.

[0018] Preferably, inputting the operating parameters into a preset time domain feature extraction model includes:

[0019] Extracting a mutation signal segment from the operating parameter, wherein the mutation signal segment is a data segment in which the parameter change rate exceeds a preset mutation rate threshold within a continuous sampling period;

[0020] generating an abnormality quantitative index based on the energy distribution and frequency domain characteristics of the mutation signal segment;

[0021] The corresponding analysis algorithm is dynamically selected according to the abnormal quantitative index, wherein the high-frequency transient mutation adopts the wavelet transform algorithm, and the low-frequency slow-changing anomaly adopts the hidden Markov model algorithm.

[0022] Preferably, the method further comprises:

[0023] After marking the abnormal signal segment, performing a signal integrity check on the operating parameters;

[0024] If the signal distortion rate is found to exceed a preset second threshold value during verification, the fault evolution model is triggered to compensate and correct the distorted signal, wherein the high-priority distorted signal is a fault data segment involving a trunk line.

[0025] Preferably, the fault evolution model includes the following prediction steps:

[0026] According to the physical topology network of the transmission line, a dynamic coupling model is constructed, in which each line node corresponds to a coupling strength coefficient;

[0027] Calculate the evolution prediction weight based on the environmental interference differences between adjacent line nodes;

[0028] Combined with the historical evolution trend of the operating parameters, multi-scale interpolation is performed to complete the missing line nodes.

[0029] Preferably, the method further comprises:

[0030] After the interpolation is completed, the prediction results are verified for physical rationality, where the verification method includes comparing the state deviation between the predicted parameters and the actual line nodes;

[0031] If the deviation exceeds a preset third threshold, the evolution prediction weight is readjusted and iteratively calculated until the deviation is less than the third threshold.

[0032] Preferably, hierarchically aggregating the prediction results and current operating parameters using a preset dynamic optimization algorithm includes:

[0033] Classify the first-level classification labels according to the fault type, wherein the first-level classification labels include emergency fault class, potential fault class and false alarm interference class;

[0034] Under each level of classification label, the secondary classification sub-labels are further divided based on the fault impact scope;

[0035] The classified fault data are stored in different partitions of the spatial database according to the label level.

[0036] Preferably, the method further comprises:

[0037] Configure access levels for classification tags based on preset operation and maintenance permissions;

[0038] Upon receiving a data query request, verify whether the permission credentials provided by the requester match the access level of the target classification label;

[0039] If there is a match, the data query interface for the corresponding classification label will be opened.

[0040] Preferably, the present invention further includes a real-time monitoring platform for power transmission lines, the platform comprising:

[0041] A multimodal sensor module, configured to collect operating parameters of the power transmission line in real time through a multimodal sensor array, wherein the operating parameters include current fluctuation value, temperature gradient distribution, mechanical vibration intensity, and insulation aging index;

[0042] a time-domain feature extraction module, configured to input the operating parameters into a preset time-domain feature extraction model to identify and mark abnormal signal segments, wherein the time-domain feature extraction model dynamically adjusts the analysis window length based on the waveform characteristics of different defect modes in historical fault data;

[0043] A fault evolution module, configured to input the marked operating parameters into a preset fault evolution model to generate a risk diffusion prediction result, wherein the fault evolution model performs a multi-scale coupling analysis based on the line topology and environmental interference factors;

[0044] The dynamic optimization module is used to perform hierarchical aggregation on the prediction results and current operating parameters using a preset dynamic optimization algorithm, generate a real-time monitoring data set and store it.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] At the data acquisition level, a multimodal sensor array enables real-time and comprehensive collection of key operating parameters of transmission lines, including current fluctuations, temperature gradient distribution, mechanical vibration intensity, and insulation aging index. This simultaneous multi-parameter acquisition overcomes the limitation of traditional monitoring methods, which only capture a single parameter, and provides a rich data foundation for accurately assessing line operating status. For example, when abnormal current fluctuations occur, combined with changes in temperature gradient distribution and mechanical vibration intensity, it is possible to more accurately determine whether the cause is a load change or a line fault, significantly improving the accuracy of fault diagnosis.

[0047] For abnormal signal identification, a preset time-domain feature extraction model dynamically adjusts the analysis window length based on the waveform characteristics of different defect patterns in historical fault data. This adaptive analysis method is more sensitive to subtle changes in abnormal signals than fixed-window analysis methods. For example, traditional fixed-window analysis may miss some early, weak fault signals due to improper window selection. However, the dynamic adjustment window of the present invention automatically adapts to the characteristics of the fault signal, improving the recognition rate of abnormal signals and ensuring that early faults are discovered in a timely manner, preventing further development of the fault.

[0048] The fault evolution model performs a multi-scale coupling analysis based on the line topology and environmental interference factors to generate risk diffusion prediction results. This model fully considers the complex factors in the actual operation of transmission lines. It can not only predict the occurrence of faults, but also accurately simulate the fault diffusion path and impact range. Taking the example of localized insulation damage caused by lightning strikes, the model can analyze the propagation of fault currents in different line branches based on the line topology. In combination with environmental interference factors (such as the surrounding electromagnetic environment and meteorological conditions), it predicts the impact of the fault on surrounding lines and equipment. This provides a scientific basis for operation and maintenance personnel to formulate targeted preventive measures in advance, effectively reducing the losses caused by the fault.

[0049] In terms of data processing and management, a preset dynamic optimization algorithm is used to hierarchically aggregate the prediction results and current operating parameters and store them as a real-time monitoring data set. By dividing the first-level classification labels according to the fault type, and then dividing the second-level classification sub-labels based on the fault impact range, and storing the data in different partitions of the spatial database according to the label level, efficient classification storage and management of data is achieved. This enables operation and maintenance personnel to quickly locate the required information when querying data, greatly improving data retrieval efficiency. At the same time, the access level of the classification label is configured according to the preset operation and maintenance permissions, which enhances the security and confidentiality of the data, ensuring that only personnel with the corresponding permissions can access key data, preventing data leakage and misoperation.

[0050] In terms of signal integrity verification and fault compensation and correction, operating parameters are verified for signal integrity. If the signal distortion rate exceeds a preset threshold, the fault evolution model is triggered to compensate for the distorted signal, especially prioritizing high-priority distorted signals involving trunk lines. This mechanism ensures the accuracy of monitoring data, avoids misjudgments or missed faults due to signal distortion, and further improves the reliability of transmission line monitoring.

[0051] Overall, the present invention has built a complete and efficient real-time monitoring system for transmission lines from data collection, analysis, prediction to data management, which effectively improves the operation and maintenance efficiency and safety of transmission lines, reduces the operation risks of power systems, and has important practical value and broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a working principle diagram of the real-time monitoring method for power transmission lines according to the present invention;

[0053] Figure 2 Workflow diagram for inputting operating parameters into the time domain feature extraction model;

[0054] Figure 3 Workflow diagram for verifying the physical rationality of prediction results;

[0055] Figure 4 Workflow diagram for data query permission verification. DETAILED DESCRIPTION

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0057] See also Figure 1-Figure 4The present invention relates to a real-time monitoring method and system for power transmission lines, and its specific implementation methods will be described in detail below.

[0058] Receive multiple read and write requests and categorize each request according to preset dimensions to generate structured request data. These dimensions include request type (e.g., read request, write request), data block size (e.g., 8KB, 16KB, etc.), priority level (high, medium, low), and storage unit location (accurately defined to the specific storage area within the memory chip). This step provides a preliminary analysis of the original requests, structuring them for easier processing.

[0059] A multi-dimensional timing constraint model is constructed, defining a timing coordinate system based on the classification dimensions of the request data. This coordinate system includes a time axis, a storage unit axis, a data flow axis, and an energy consumption axis. The time axis measures the order and time intervals of operations; the storage unit axis corresponds to the storage unit layout of the memory chip; the data flow axis reflects the direction and path of data transmission within the memory chip; and the energy consumption axis focuses on the energy consumption of the memory chip during the read and write processes. This coordinate system enables comprehensive consideration and constraints on read and write tasks from multiple dimensions.

[0060] Initial timing constraints are determined based on the request type and data block size. These constraints include read and write latency thresholds (specifying the maximum allowable latency for read or write operations), data path conflict rules (defining the rules that prevent conflicts between different requests on the data transmission path), and resource allocation priorities (determining the order in which different requests receive resources when resources are limited). These constraints provide the foundation for subsequent scheduling and optimization.

[0061] Based on the timing constraint model, data flow simulation is performed on read and write tasks. A genetic algorithm is used to simulate the dynamic distribution of data within storage units to generate a preliminary scheduling solution. By simulating natural evolutionary processes such as selection, crossover, and mutation, the genetic algorithm continuously optimizes the distribution of data within storage units, resulting in a preliminary read and write task scheduling solution.

[0062] Path conflict detection is performed on the preliminary scheduling plan to identify storage nodes with overlapping data paths or rule conflicts. The conflict type (for example, data path overlap or priority rule violation) and latency intensity (indicating the impact of the conflict on system performance) are marked. This step can identify problems in the preliminary scheduling plan and provide a basis for subsequent optimization.

[0063] A dynamic priority scheduling algorithm is used to iteratively optimize conflicting nodes. This algorithm adjusts storage node timing parameters (such as operation start time and duration) or resource allocation strategies (such as changing resource allocation priority and allocation quantity) to generate a conflict-free optimized scheduling solution. Through continuous iterative optimization, a conflict-free scheduling solution that meets system performance requirements is ultimately achieved, improving the read and write efficiency and performance of the storage chip.

[0064] The present invention will be further described below in conjunction with Examples 1 to 5:

[0065] Example 1:

[0066] This embodiment details a specific application scenario of a memory chip read / write control method. In an actual storage system, a memory chip continuously receives read / write requests from various devices. For example, in a server storage system, multiple applications are running simultaneously, frequently performing data read / write operations on the memory chip.

[0067] Assume that a storage chip receives a series of read and write requests, including database read requests and file write requests. The data block sizes vary, such as 4KB and 16KB. The priority levels are high for urgent tasks, medium for ordinary tasks, and low for background tasks. The storage unit locations are located in different storage areas within the storage chip, such as Area A and Area B.

[0068] According to claim 2, after obtaining the optimized scheduling scheme, it is necessary to synchronize the optimized scheduling scheme with the real-time load data of the storage chip and dynamically correct the parameters of the timing constraint model. The real-time load data of the storage chip includes the current utilization rate of the storage unit, the current read and write speed, etc. Assuming that the current storage unit utilization rate is U and the current read and write speed is S, through the formula T new =T old ×(1+k1×(UU avg ))×(1+k2×(SS avg )) Modify the time axis related parameters, including T new is the corrected time axis parameter value, T old is the time axis parameter value before correction, U avg is the average usage rate of storage units, S avg is the average value of the read and write speeds, and k1 and k2 are weight coefficients set according to the actual system conditions.

[0069] At the same time, based on the revised timing constraint model, the final scheduling instruction set is generated. The instruction set contains read and write operation timing, data path planning, and resource allocation information. For example, for a high-priority database read request, the instruction set will clearly specify when to start the read operation, which storage unit the data is transmitted from and which data path, and how many system resources are allocated for the read operation. Afterwards, the scheduling instruction set is pushed to the storage controller and execution unit, triggering the automated scheduling process. The storage controller controls the storage chip to perform the corresponding read and write operations according to the instruction set, and the execution unit transmits and processes data according to the instruction requirements, thereby achieving efficient and orderly storage chip read and write control.

[0070] Example 2:

[0071] This example uses a common solid-state drive (SSD) memory chip as an example to illustrate the process of building a multidimensional timing constraint model. This SSD memory chip is widely used for data storage in personal computers and processes a large number of read and write requests from operating systems and various applications every day.

[0072] In terms of storage unit axis partitioning, the memory chip's internal structure allows for three main partitions. Partition 1 is located near the memory chip's control core area, which provides high access bandwidth and enables fast data transfer. This makes it suitable for storing critical system data and frequently read / written application data. For example, operating system boot files and commonly used office software data can be stored in this partition to speed up system startup and software loading. This partition also features a large cache capacity for temporary storage of frequently accessed data, reducing direct accesses to the storage unit and improving read / write efficiency.

[0073] Partition 2 is located in the middle of the memory chip, offering moderate access bandwidth and a relatively balanced physical location. This partition can be used to store user-generated data, such as documents and images, which requires relatively low read and write frequency but still requires a certain level of access speed. Its cache capacity is slightly smaller than partition 1, balancing cost and performance while meeting daily needs.

[0074] Partition 3 is located at the edge of the storage chip, offering relatively low access bandwidth but also relatively low cost. This partition is suitable for storing infrequently accessed cold data, such as long-term backup files and historical records. Because this data is rarely used and requires low access speed, storing it here maximizes the storage chip's space without significantly impacting overall performance.

[0075] In terms of timeline division, it is set as a discrete scheduling cycle, each cycle is 10 milliseconds. In different time periods, the system's processing requirements for requests of different priorities will vary. For example, when the computer is just turned on, the system will give priority to high-priority read requests related to the startup of the operating system. At this time, the weight of high-priority requests is set higher. As the system enters a stable operating state, the read and write requests of various applications increase, and the weight of medium-priority requests will be adjusted according to actual conditions. When the computer is idle, low-priority background data backup requests will be processed, and their weight will be increased accordingly. By dynamically adjusting the weights of requests of different priorities in each scheduling cycle, system resources can be allocated more reasonably to ensure that important tasks are executed first, while not neglecting the processing of other tasks.

[0076] Regarding the energy consumption axis, memory chips consume energy during read / write operations, standby mode, and heat dissipation. During read / write operations, energy consumption varies depending on the amount of data and the required read / write speed. For example, high-speed writes of large amounts of data significantly increase read / write power consumption. When the memory chip is in standby mode (i.e., when no read / write requests are being made), it still consumes a certain amount of standby power to maintain the data storage state of the memory cells and basic system operation. Heat dissipation energy consumption is closely related to the chip's operating temperature. When the chip operates at high load for extended periods, causing the temperature to rise, the heat dissipation system increases its workload to ensure chip stability and performance, thereby increasing heat dissipation energy consumption. By dynamically managing these three energy consumption states, for example, when the chip temperature is low, read / write speeds are appropriately increased to improve efficiency while controlling heat dissipation energy consumption; when the temperature is too high, read / write speeds are reduced and heat dissipation energy consumption is increased. This ensures that the chip always operates within the appropriate temperature range, effectively allocating energy consumption to the memory chip, extending its lifespan, and reducing overall energy consumption.

[0077] Example 3:

[0078] This example uses a high-performance storage chip for big data storage as an example to illustrate the practical application of a dynamic priority scheduling algorithm. In big data storage scenarios, the storage chip simultaneously receives requests from multiple data analysis and data write tasks. These requests vary in size and priority, making them prone to conflicting nodes. This requires optimization using a dynamic priority scheduling algorithm.

[0079] Suppose, at a certain moment, a memory chip receives 10 read and write requests. After preliminary scheduling, it is discovered that three of these requests conflict in data transmission paths and resource allocation, indicating the presence of conflicting nodes. The dynamic priority scheduling algorithm uses an improved weighted round-robin mechanism, using the timing parameters of conflicting nodes as key decision-making criteria. For example, timing parameters such as the start time and duration of conflicting nodes directly influence the execution order of requests and resource usage.

[0080] An objective function is defined to comprehensively evaluate latency intensity, resource utilization, and scheduling deviation cost. Latency intensity primarily measures the impact of conflicts on read and write latency. If a high-priority request is delayed due to a conflict, the latency intensity will be higher, and the impact on system performance will be more severe. Resource utilization focuses on whether the various resources of the memory chip, such as storage units and data transmission channels, are fully and reasonably utilized. Scheduling deviation cost reflects the gap between the actual scheduling plan and the ideal scheduling plan. The larger the gap, the higher the scheduling deviation cost.

[0081] In practice, multiple new scheduling schemes are first generated based on the conflicting nodes. For example, by adjusting the execution order of conflicting requests or allocating different storage resources, a series of possible scheduling schemes are obtained. Each scheme is then evaluated, calculating its latency intensity, resource utilization, and scheduling deviation cost. For example, a scheme that allows high-priority requests to execute as quickly as possible, reduces latency intensity, improves resource utilization, and reduces scheduling deviation cost is more likely to be retained.

[0082] A weight adjustment strategy is used to retain the optimal solution set. Initially, weights are set for latency intensity, resource utilization, and scheduling deviation cost based on the system's performance requirements and application scenario characteristics. For example, in real-time data analysis scenarios with high latency requirements, a higher weight is assigned to latency intensity; in large-scale data storage scenarios that are more sensitive to resource costs, a higher weight is assigned to resource utilization. After evaluating all newly generated scheduling plans, a comprehensive score is calculated for each plan based on the weights, and the plan with the highest score is retained as the optimal solution set.

[0083] As the system operates and new conflicts arise, weights are dynamically adjusted. If the system is experiencing frequent high latency issues, the weight of latency intensity is appropriately increased, and the scheduling plan is re-evaluated and screened. Through continuous iteration of this process, the timing parameters and resource allocation strategies of storage nodes are gradually adjusted, ultimately generating a conflict-free, optimized scheduling plan. This ensures that the storage chips can efficiently and stably handle various read and write requests in complex big data storage environments, improving the performance and reliability of the entire storage system.

[0084] Example 4:

[0085] This embodiment describes in detail the specific implementation process of data flow simulation. In the memory chip read and write control method, data flow simulation is an important step in generating a reasonable scheduling solution.

[0086] For example, consider a storage system with real-time data processing requirements, such as a video surveillance data storage system. In this system, the storage chip needs to continuously receive video data write requests from cameras, while other devices may also request to read the stored video data.

[0087] During the data flow simulation, a real-time load monitoring module is introduced to dynamically receive data on the access frequency of the storage unit, the cache hit rate, and the request queue length. Assume that the access frequency of the storage unit is f, the cache hit rate is h, and the request queue length is q. The boundary conditions of the simulation model are adjusted according to the real-time data. For example, when the access frequency of the storage unit f increases, the data transmission bandwidth is appropriately increased to avoid data congestion. By using formula B new =B old ×(1+k×f) adjusts the data transmission bandwidth B in the data flow axis, where B new is the adjusted bandwidth value, B old is the bandwidth value before adjustment, and k is the adjustment coefficient set according to system performance.

[0088] When the cache hit rate h is low, it means that the cache utilization efficiency is not high, and it may be necessary to adjust the storage strategy of data in the storage unit. <h threshold (h threshold If the cache hit rate threshold is set, a more reasonable cache replacement algorithm, such as an improved version of the least recently used (LRU) algorithm, is used to prioritize replacing infrequently used data from the cache, freeing up space for new data and improving the cache hit rate. Simultaneously, the scheduling cycle in the simulation model is adjusted based on the request queue length q. When q is long, the scheduling cycle is appropriately shortened to speed up request processing and avoid long request backlogs. By comprehensively considering this real-time data and adjusting the boundary conditions of the simulation model, a fault-tolerant scheduling scheme is generated, enabling the memory chip to better cope with various complex workloads and improving the reliability and efficiency of data reading and writing.

[0089] Example 5:

[0090] This embodiment elaborates on path conflict detection and data path planning. During the reading and writing process of the memory chip, path conflict detection and reasonable data path planning are crucial to ensure normal operation of the system and improve performance.

[0091] In terms of path conflict detection, we take a storage system containing multiple storage modules as an example. We establish a data path occupancy matrix. Assuming that the storage system has m storage nodes, n scheduling cycles, and p storage unit partitions, the data path occupancy matrix M is a three-dimensional matrix of m×n×p. The matrix element M i,j,k Indicates the path occupancy of the i-th storage node in the j-th scheduling cycle and the k-th storage unit partition. When M i,j,k =1, indicating that the storage node is occupied, M i,j,k =0 indicates unoccupied. Matrix operations are used to detect the overlapping areas of storage nodes in scheduling cycles and storage unit partitions. For example, the sum of matrix elements of different storage nodes in the same scheduling cycle and storage unit partition is calculated. If the sum is greater than 1, it indicates that there is a path overlap conflict.

[0092] The rules engine matches pre-set conflict determination logic to identify access priority or bandwidth limit violations. Assume there is an access priority rule: high-priority requests have priority over low-priority requests in accessing storage resources. If a low-priority request accesses the same storage resource before a high-priority request, it is considered an access priority violation.

[0093] In terms of data path planning, a fuzzy logic control algorithm is used. The storage channel topology is modeled as a state transition network, and the data flow diffusion action is defined as a fuzzy state. Assume that the storage channel has multiple nodes such as A, B, and C, and the action of data flow diffusion from node A to node B is a fuzzy state S AB . Path selection is guided by the membership update strategy to minimize access delay and avoid local congestion. Let the access delay be d and the congestion level be c, define the membership functions μ(d) and μ(c), and calculate the membership value based on the current access delay and congestion level. For example, when the access delay d is small and the congestion level c is low, the membership value of the selected path is higher. By continuously updating the membership value, the data flow is guided to select the optimal data path, improving the data transmission efficiency of the storage chip, reducing the occurrence of conflicts, and ensuring the stable operation of the storage system.

[0094] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0095] While 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 these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A real-time monitoring method for a transmission line, characterized in that: include: Real-time acquisition of transmission line operating parameters using a multimodal sensor array, wherein the operating parameters include current fluctuation value, temperature gradient distribution, mechanical vibration intensity, and insulation aging index; Inputting the operating parameters into a preset time-domain feature extraction model to identify and mark abnormal signal segments, wherein the time-domain feature extraction model dynamically adjusts the analysis window length based on the waveform characteristics of different defect modes in historical fault data; Inputting the marked operating parameters into a preset fault evolution model to generate risk diffusion prediction results, wherein the fault evolution model performs multi-scale coupling analysis based on the line topology and environmental interference factors; The prediction results and current operating parameters are hierarchically aggregated using a preset dynamic optimization algorithm to generate and store a real-time monitoring data set.

2. The real-time monitoring method for power transmission lines according to claim 1, characterized in that: The steps of constructing the time domain feature extraction model include: obtaining a historical fault data set, wherein each data in the historical fault data set is marked with a defect type and a hazard level; dividing training subsets based on the defect type and hazard level, each training subset corresponding to a fault scenario; using the training subsets to train the initial feature extraction model in parallel, until the waveform recognition accuracy of the initial feature extraction model for each fault scenario is greater than or equal to a preset first threshold, stopping the training, and obtaining an intermediate feature extraction model; inputting the historical fault data set into the intermediate feature extraction model, and verifying whether the feature matching degree output by the intermediate feature extraction model meets the preset error range; if so, determining the intermediate feature extraction model as the time domain feature extraction model.

3. The real-time monitoring method for power transmission lines according to claim 1, characterized in that: The real-time acquisition of transmission line operating parameters by a multimodal sensor array includes: Establishing a communication link with a distributed monitoring node, wherein the distributed monitoring node is deployed at a preset key monitoring location of the transmission line; Continuously acquiring the real-time signal of the distributed monitoring node according to a preset sampling frequency, and marking an acquisition time stamp based on the spectrum characteristics of the real-time signal; According to the physical connection relationship of the transmission lines, the real-time signals of different monitoring locations at the same timestamp are spatially aligned to form a set of associated operating parameters.

4. The real-time monitoring method for transmission lines according to claim 1, characterized in that: Inputting the operating parameters into a preset time domain feature extraction model includes: Extracting a mutation signal segment from the operating parameter, wherein the mutation signal segment is a data segment in which the parameter change rate exceeds a preset mutation rate threshold within a continuous sampling period; generating an abnormality quantitative index based on the energy distribution and frequency domain characteristics of the mutation signal segment; The corresponding analysis algorithm is dynamically selected according to the abnormal quantitative index, wherein the high-frequency transient mutation adopts the wavelet transform algorithm, and the low-frequency slow-changing anomaly adopts the hidden Markov model algorithm.

5. The real-time monitoring method for power transmission lines according to claim 4, characterized in that: The method further comprises: After marking the abnormal signal segment, performing a signal integrity check on the operating parameters; If the signal distortion rate is found to exceed a preset second threshold value during verification, the fault evolution model is triggered to compensate and correct the distorted signal, wherein the high-priority distorted signal is a fault data segment involving a trunk line.

6. The method for real-time monitoring of a power transmission line according to claim 1, wherein: The fault evolution model includes the following prediction steps: According to the physical topology network of the transmission line, a dynamic coupling model is constructed, in which each line node corresponds to a coupling strength coefficient; Calculate the evolution prediction weight based on the environmental interference differences between adjacent line nodes; Combined with the historical evolution trend of the operating parameters, multi-scale interpolation is performed to complete the missing line nodes.

7. The real-time monitoring method for power transmission lines according to claim 6, characterized in that: The method further comprises: After the interpolation is completed, the prediction results are verified for physical rationality, where the verification method includes comparing the state deviation between the predicted parameters and the actual line nodes; If the deviation exceeds a preset third threshold, the evolution prediction weight is readjusted and iteratively calculated until the deviation is less than the third threshold.

8. The real-time monitoring method for power transmission lines according to claim 1, characterized in that: The hierarchical aggregation of the prediction results and current operating parameters using a preset dynamic optimization algorithm includes: Classify the first-level classification labels according to the fault type, wherein the first-level classification labels include emergency fault class, potential fault class and false alarm interference class; Under each level of classification label, the secondary classification sub-labels are further divided based on the fault impact scope; The classified fault data are stored in different partitions of the spatial database according to the label level.

9. The method for real-time monitoring of a power transmission line according to claim 8, characterized in that: The method further comprises: Configure access levels for classification tags based on preset operation and maintenance permissions; Upon receiving a data query request, verify whether the permission credentials provided by the requester match the access level of the target classification label; If there is a match, the data query interface for the corresponding classification label will be opened.

10. A real-time monitoring platform for power transmission lines, characterized in that: include: A multimodal sensor module, configured to collect operating parameters of the power transmission line in real time through a multimodal sensor array, wherein the operating parameters include current fluctuation value, temperature gradient distribution, mechanical vibration intensity, and insulation aging index; a time-domain feature extraction module, configured to input the operating parameters into a preset time-domain feature extraction model to identify and mark abnormal signal segments, wherein the time-domain feature extraction model dynamically adjusts the analysis window length based on the waveform characteristics of different defect modes in historical fault data; A fault evolution module, configured to input the marked operating parameters into a preset fault evolution model to generate a risk diffusion prediction result, wherein the fault evolution model performs a multi-scale coupling analysis based on the line topology and environmental interference factors; The dynamic optimization module is used to perform hierarchical aggregation on the prediction results and current operating parameters using a preset dynamic optimization algorithm, generate a real-time monitoring data set and store it.

Images