Power supply data adaptive transmission optimization method, system, equipment and medium

Abstract

The invention discloses a power supply data self-adaptive transmission optimization method, system and device and a medium. The method comprises the steps that a data fusion and service inference model is built according to the type of power supply equipment and the priority of a service link where the power supply equipment is located; a dynamic data value weight matrix is obtained through model output parameters, and value weights of the evaluation features are quantified; and adaptively adjusting a sampling strategy, a data transmission strategy and a feedback mechanism, dynamically adjusting the data value dynamic matrix and issuing an acquisition and transmission strategy. According to the invention, a sampling strategy is dynamically adjusted to preferentially guarantee high-value data transmission; optimizing the data transmission efficiency by adopting a hierarchical data transmission strategy; iteratively updating the data value dynamic matrix through a cloud edge collaborative feedback mechanism, and issuing an acquisition transmission strategy to continuously optimize the data transmission efficiency; multi-source data fusion is utilized to construct a data weight dynamic matrix, model output, business rules and timeliness factors are comprehensively considered, and the data value is comprehensively evaluated.

Description

Technical Field

[0001] The present invention relates to the field of data acquisition optimization, and particularly to a power supply data adaptive transmission optimization method, system, device and medium. Background Art

[0002] In recent years, with the accelerated intelligent transformation of the distribution network, the scale of data generated by a large number of terminal devices has increased exponentially, which will occupy a large amount of network bandwidth and storage resources, resulting in low data transmission efficiency and inability to meet real-time requirements. Existing data transmission solutions are insufficient in data value mining, resource dynamic allocation, and cloud-edge collaborative analysis. The main manifestations are that the existing technology lacks the dynamic evaluation ability of multi-dimensional fusion for the value and feature fusion of power supply data. The data interaction between the edge side and the cloud side is still mainly one-way transmission, adopting the method of centralized processing in the cloud. The edge side devices lack data processing and analysis capabilities and are difficult to perform dynamic adjustment through cloud-edge collaboration. Existing power supply data acquisition and transmission mostly adopt fixed frequency and full-volume transmission, resulting in the loss and undetected of transient features and the occupation of bandwidth by low-value data.

[0003] Therefore, the existing power supply data acquisition and transmission strategy has low resource utilization rate and insufficient capture of key information. There is an urgent need to develop a data acquisition and transmission optimization method with scene adaptation to improve the real-time monitoring and transmission ability of power supply quality and provide data support for power supply situation awareness and fault warning. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by the present invention are: the existing data transmission optimization methods have low data transmission efficiency, insufficient capture of key information, insufficient data value mining, insufficient cloud-edge collaboration ability, and the problem of how to achieve data adaptive transmission optimization.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a power supply data adaptive transmission optimization method, including:

[0008] Construct a data fusion and service reasoning model according to the power supply device type and the priority of the business link where it is located; obtain a dynamic data value weight matrix through the output parameters of the data fusion and service reasoning model, and quantitatively evaluate the value weight of features; adaptively adjust the sampling strategy according to the value weight of features, and dynamically select a data transmission strategy according to the communication channel condition; dynamically adjust the data value dynamic matrix according to the change of the feature contribution value fed back by the feedback mechanism and issue a collection and transmission strategy.

[0009] As a preferred solution of the power supply data adaptive transmission optimization method described in the present invention, wherein: the value weights of the quantization evaluation features include:

[0010] Normalize the event probabilities and business process feature data parameters output by the data fusion and business inference models; set the business data correlation degree through the business rule library and convert it into a numerical weight coefficient; calculate the feature timeliness coefficient based on the time sliding window, and evaluate the value of historical data through the time decay function; perform multi-source data fusion through the data analysis method, and evaluate the model output parameters, business rules, and feature timeliness coefficients; determine the weight allocation of each dimension, construct a dynamic data value weight matrix, and calculate the value weights of the features.

[0011] As a preferred solution of the power supply data adaptive transmission optimization method described in the present invention, wherein: the adaptive adjustment of the sampling strategy according to the value weights of the features includes:

[0012] Adopt a dynamic window sampling strategy for the time series features of high-value weight features, automatically adjust the sampling window with the change gradient of the features, and adjust different sampling modes according to the power supply business scenario; adopt a sparse sampling strategy for low-value weight features.

[0013] As a preferred solution of the power supply data adaptive transmission optimization method described in the present invention, wherein: the dynamic selection of the data transmission strategy according to the communication channel conditions includes:

[0014] Deploy a circular storage buffer on the power supply acquisition terminal side to continuously overwrite the original power supply operation data; real-time detect the communication channel conditions, and divide the channel conditions into high, medium, and low grades according to the communication frequency band and transmission bandwidth; adopt a hierarchical data transmission strategy according to the communication channel grades.

[0015] As a preferred solution of the power supply data adaptive transmission optimization method described in the present invention, wherein: the change of the feature contribution value feedback by the feedback mechanism includes:

[0016] Build a data fusion model on the edge side, dynamically select the data transmission strategy, upload data on the terminal side and input it into the data fusion model; the data fusion model returns the importance of data features, dynamically adjust the data value dynamic matrix through the change of the feature contribution value feedback by cloud-edge collaboration, and iteratively optimize the data feature acquisition and transmission strategy.

[0017] As a preferred solution of the power supply data adaptive transmission optimization method described in the present invention, wherein: the sampling modes include a fault-sensitive mode, a power quality mode, and an economic operation mode.

[0018] As a preferred solution of the power supply data adaptive transmission optimization method described in the present invention, wherein: the hierarchical data transmission strategy includes:

[0019] When the terminal side is in a high - bandwidth environment, an improved compression algorithm is used to transmit data losslessly; when the terminal side is in a medium - bandwidth environment, a segmented compression transmission strategy is adopted; when the terminal side is in a low - bandwidth environment, a feature screening strategy is adopted.

[0020] In a second aspect, the present invention provides a power supply data adaptive transmission optimization system, including:

[0021] A feature evaluation module, a data sampling and transmission module, and an update and optimization module;

[0022] The feature evaluation module is used to obtain a dynamic data value weight matrix through the output parameters of the data fusion and service inference model, and quantitatively evaluate the value weight of the feature;

[0023] The data sampling and transmission module is used to adaptively adjust the sampling strategy according to the value weight of the feature; dynamically select the data transmission strategy according to the communication channel conditions;

[0024] The update and optimization module is used to dynamically adjust the data value dynamic matrix according to the change of the feature contribution value feedback by the feedback mechanism and issue the acquisition and transmission strategy.

[0025] In a third aspect, the present invention provides an electronic device, including:

[0026] A memory and a processor;

[0027] The memory is used to store computer - executable instructions, and the processor is used to execute the computer - executable instructions. When the computer - executable instructions are executed by the processor, the steps of the power supply data adaptive transmission optimization method are implemented.

[0028] In a fourth aspect, the present invention provides a computer - readable storage medium, which stores computer - executable instructions. When the computer - executable instructions are executed by a processor, the steps of the power supply data adaptive transmission optimization method are implemented.

[0029] Compared with the prior art, the beneficial effects of the present invention are as follows: By calculating data value and feature weight, the present invention dynamically adjusts the sampling strategy, avoiding low-value data from occupying bandwidth and preferentially ensuring the transmission of high-value data; According to the communication channel conditions, a hierarchical data transmission strategy is adopted to optimize data transmission efficiency; Through the cloud-edge collaborative feedback mechanism, according to the changes in data contribution degree and feature sensitivity, the data value dynamic matrix is iteratively updated and the acquisition and transmission strategy is issued to continuously optimize data transmission efficiency; The dynamic matrix of data weights is constructed by using multi-source data fusion, comprehensively considering model output, business rules, and timeliness factors, and comprehensively evaluating data value; Through the adaptive sampling strategy, aiming at different power supply scenarios and business execution differences, the priority acquisition and transmission of key data and high-value features are ensured; By constructing a data fusion model on the edge side and dynamically selecting a hierarchical transmission strategy, the ability to capture key features can be further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0031] Figure 1 It is a schematic diagram of the overall process of the power supply data adaptive transmission optimization method according to an embodiment of the present invention.

[0032] Figure 2 It is a schematic diagram of the detailed judgment process of the power supply data adaptive transmission optimization method according to an embodiment of the present invention.

[0033] Figure 3 It is a schematic diagram of the overall process of the power supply data adaptive transmission optimization method according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following detailed description of the specific embodiments of the present invention will be made in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0035] Embodiment 1, referring to Figure 1 , which is an embodiment of the present invention, provides a power supply data adaptive transmission optimization method, including:

[0036] S1: Construct a data fusion and business reasoning model based on the power supply equipment type and the priority of the business link where it is located;

[0037] S3: Obtain a dynamic data value weight matrix through the output parameters of the data fusion and business reasoning model, and quantitatively evaluate the value weights of the features;

[0038] S3: Adaptively adjust the sampling strategy according to the value weights of the features, and dynamically select the data transmission strategy according to the communication channel conditions;

[0039] S4: Dynamically adjust the data value dynamic matrix according to the change of the feature contribution value feedback by the feedback mechanism, and issue the acquisition and transmission strategy.

[0040] It should be noted that when power supply equipment operates in the power system, a large amount of data is often generated, such as equipment status, load conditions, power quality, etc. These data need to be transmitted to the control center in real time for processing in order to monitor and analyze the status of the power supply equipment, and timely discover and handle potential fault hazards. Therefore, it is very important to optimize the data transmission method and achieve efficient real-time transmission of data in the power system.

[0041] Therefore, in response to the above-mentioned data transmission problems, through steps S1-S4, a data fusion and business reasoning model is constructed, a dynamic data value weight matrix is obtained, and the value weights of the features are quantitatively evaluated, realizing the multi-dimensional fusion dynamic evaluation ability lacking in the prior art for the value and feature fusion of power supply data; and through the adaptive sampling strategy and dynamic transmission strategy, the quality of key data and the utilization rate of processing and transmission resources are improved, realizing the accurate capture of power supply data features and enhancing the power supply quality level; through the cloud-edge collaborative feedback mechanism, two-way interaction of data between the edge side and the cloud side is realized, and a data fusion model is constructed on the edge side to improve the data processing and analysis ability on the edge side. At the same time, through the feedback mechanism, the data value weight matrix is iteratively optimized and the acquisition strategy is issued to continuously optimize the data transmission efficiency.

[0042] Example 2, refer to Figure 2 , which is an embodiment of the present invention. Based on the above embodiment, an adaptive transmission optimization method for power supply data is provided.

[0043] In the implementation manner of the present application, for the power supply equipment type and the priority of the business link where it is located obtained by constructing the data fusion and business reasoning model in step S1, machine learning models such as data fusion algorithms and decision trees are used to perform data fusion and priority reasoning on the power supply equipment type data and business priorities.

[0044] In the implementation manner of the present application, the quantitative evaluation of the value weights of the features in step S2 includes the following steps A1-A5:

[0045] A1: Normalize the event probabilities and business process feature data parameters output by the data fusion and business inference models;

[0046] A2: Set the business data correlation degree through the business rule library and convert it into a numerical weight coefficient;

[0047] A3: Calculate the feature timeliness coefficient based on the time sliding window and evaluate the value of historical data through the time decay function;

[0048] A4: Perform multi-source data fusion through data analysis methods and evaluate the model output parameters, business rules, and feature timeliness coefficients;

[0049] A5: Determine the weight allocation for each dimension, construct a dynamic data value weight matrix, and calculate the value weight of the features.

[0050] Specifically, in step A1, directly map the event probabilities output by the data fusion and business inference models to [0, 1], denoted as U, and use the softmax function to normalize and map the business process-related feature data to the same interval, denoted as F. The data value is jointly determined by U and F. The calculation formula for the data value of the model is expressed as:

[0051] S m = w1 * U + w2 * F

[0052] Among them, w1 and w2 are the weight coefficients of the event probability and business process feature data respectively;

[0053] Among them, w1 is obtained from expert experience, and w2 is related to the feature contribution value feedback from the cloud.

[0054] In step A2, set the business rule library based on expert experience and operation and maintenance practices according to the distribution network equipment type and power supply reliability. The specifically set business rule library includes that if the equipment of the key power supply node has been running for more than 8 years and the load rate is greater than 85% and continues for more than 1 hour, the abnormal probability is relatively high, and the impact intensity is set to level one; if the general load equipment meets the conditions of running for 3 - 8 years and the load rate is less than 85%, the abnormal probability is medium, and the impact intensity is set to level two; if the equipment has been running for less than 3 years, the abnormal probability is relatively low, and the impact intensity is set to level three; among them, when the equipment failure rate is greater than 5% or the energy efficiency is lower than the standard value, the impact intensity is increased by one level, with the highest being level one; when it is less than 1%, the impact intensity is decreased by one level;

[0055] Set the business rule weights. The weight coefficient for level one is 0.8, the weight coefficient for level two is 0.5, and the weight coefficient for level three is 0.2.

[0056] In step A3, set the time decay function, expressed as:

[0057]

[0058] where \(t\) is the interval between the data generation time and the current time, and \(T\) max is the maximum time span of the sliding window, and \(\lambda\) is the attenuation coefficient;

[0059] Deploying a larger value of the attenuation coefficient during high - failure periods can slow down the attenuation rate of data value, and deploying a smaller value of the attenuation coefficient during steady - state operation periods can accelerate the elimination of old data.

[0060] In steps A4 and A5, a dynamic data value weight matrix \(A\) is constructed for the model, time, and rules, expressed as:

[0061]

[0062] where \(S\) r represents the business rule weight, and \(S\) m represents the data value importance score output by the data fusion and business reasoning model;

[0063] Calculate the eigenvector \(W\) of the dynamic data value weight matrix \(A\), expressed as:

[0064] \(W = [w\) m , \(w\) r , \(w\) t \) T

[0065] where \(T\) is the time span of the sliding window;

[0066] Weighted fusion of data in each dimension is performed according to the eigenvector \(W\) to generate a comprehensive data value score, and the data value score is expressed as:

[0067] \(DVM=w\) m *\(S\) m +\(w\) r *\(S\) r +\(w\) t *\(S\) t

[0068] In an alternative embodiment, the multi - source data fusion in step S2 can also automatically learn the correlation between the features of data sources in different dimensions and perform fusion through a neural network structure, such as a deep learning model.

[0069] In another alternative embodiment, step S2 can also extract representative features from data in different dimensions through feature extraction algorithms, such as principal component analysis or linear discriminant analysis, to reduce redundant information.

[0070] In the embodiment of the present application, in step S3, the sampling strategy is adaptively adjusted according to the value weight of the feature, including adopting a dynamic window sampling strategy for the time series features of high value weight features, automatically adjusting the sampling window along with the feature change gradient, and adjusting different sampling modes according to the power supply service scenario; adopting a sparse sampling strategy for low value weight features.

[0071] The sampling modes include a fault-sensitive mode, a power quality mode, and an economic operation mode;

[0072] Specifically, for high value weight features, the adjustment rule of the dynamic window sampling strategy is expressed as:

[0073]

[0074] where T b is the reference sampling period, α is the adjustment coefficient, k is the attenuation factor, and Δy / Δt is the harmonic distortion rate;

[0075] In the sampling mode, for the power supply service scenario, when the number of faults within 24 hours exceeds 1 or the voltage sag amplitude is greater than 15%, the fault-sensitive mode is triggered, and full-waveform capture of features such as voltage and current is performed, and the acquisition window is narrowed;

[0076] When the harmonic distortion rate exceeds 5%, the power quality mode is started, and synchronous phase sampling is performed on the 2nd to 25th harmonic components;

[0077] When the load rate of the transformer substation area is below 40% and lasts for 2 hours, the economic operation mode is started, and sliding average filtering is adopted for steady-state features such as current and voltage, and the sampling window is extended;

[0078] For low value weight features, a hierarchical sparse sampling strategy is implemented according to the power supply scenario. The sampling period is set to 30 minutes during the night low valley period and 5 minutes during the light load steady operation.

[0079] In the embodiment of the present application, in step S3, the data transmission strategy is dynamically selected according to the communication channel conditions, including the following steps C1 - C3:

[0080] C1: Deploy a circular storage buffer on the power supply acquisition terminal side to continuously overwrite the original power supply operation data;

[0081] C2: Real-time detect the communication channel conditions, and classify the channel conditions into high, medium, and low grades according to the communication frequency band and transmission bandwidth;

[0082] C3: Adopt a hierarchical data transmission strategy according to the communication channel grade.

[0083] The layered data transmission strategy includes adopting an improved compression algorithm to transmit data losslessly when the terminal side is in a high-end bandwidth environment; adopting a segmented compression transmission strategy when the terminal side is in a medium-end bandwidth environment; and adopting a feature screening strategy when the terminal side is in a low-end bandwidth environment.

[0084] Specifically, in step C1, continuously overwriting the original power supply operation data is to continuously overwrite and write the original power supply operation data of the latest 24 hours according to the first-in-first-out principle;

[0085] Deploy a storage buffer to divide the physical memory into N equal-length data blocks, overwrite old data through a write pointer loop, extract data according to the value weight priority of data features through a read pointer, and attach metadata tags to each data block, such as timestamp, value weight, exception mark and other information. When the buffer is full, the earliest written data block is overwritten first.

[0086] In steps C2 and C3, if the hierarchical data transmission strategy is at high level, the original data of power supply operation, full-featured time series data and abnormal events are recorded and packaged and uploaded to the cloud data center through the improved compression algorithm to ensure data integrity and real-time performance;

[0087] If the hierarchical data transmission strategy is in the middle range, the full waveform is divided into complete cycle segments with the zero crossing point as the boundary through the segmented compression transmission strategy, and only the segments with abnormal marks are retained, such as harmonic exceeding the standard, voltage exceeding the limit, etc., and the feature vectors generated by the mean, variance, extreme value, etc. in the segment are compressed and transmitted using the combined compression algorithm;

[0088] If the layered data transmission strategy is inadequate, the feature screening strategy is used to remove feature data with a value weight lower than 0.2 and then transmit it through segmented compression to ensure that key data is uploaded first, with a compression rate of 30%-50%.

[0089] In an optional implementation, the hierarchical data transmission strategy in step S3 may also use a lossless compression algorithm, such as Huffman coding and Lempel-Ziv coding, at a high level to achieve compressed transmission of data, thereby reducing the amount of data transmission without losing data information.

[0090] In another optional implementation, the hierarchical data transmission strategy in step S3 can also be based on the feature importance sorting transmission strategy when in the middle range, sorting the data by feature importance, giving priority to transmitting features with high importance, reducing the amount of transmitted data while ensuring the transmission of key information.

[0091] In the implementation manner of the present application, the characteristic contribution value change fed back by the feedback mechanism in step S4 includes the following steps D1-D2:

[0092] D1: Build a data fusion model on the edge side, dynamically select a data transmission strategy, upload data on the terminal side and input it into the data fusion model;

[0093] D2: The data fusion model returns the importance of data features, dynamically adjusts the data value dynamic matrix through the change of the feature contribution value feedback by cloud-edge collaboration, and iteratively optimizes the data feature acquisition and transmission strategy.

[0094] It should be noted that if the feature contribution value is less than the threshold within a continuous period, the weight will be automatically reduced. For long-term stable features, the weight will be frozen and the acquisition and transmission frequency will be reduced to continuously optimize the data transmission efficiency and model fusion performance.

[0095] Specifically, the feature contribution change amount ΔP i , and the calculation formula is expressed as:

[0096]

[0097] Among them, is the initial importance of the i-th feature received on the edge side in the k-th iteration, is the importance of the i-th feature feedback after the (k + 1)-th iteration;

[0098] Dynamically adjust the feature weight coefficient through the feature contribution change amount It is expressed as:

[0099]

[0100] If the feature contribution change amount ΔP i is lower than 0.05 for 12 consecutive months, the weight will be reduced to 0.2 and the acquisition will stop;

[0101] Reduce the sampling frequency for long-term stable features to further optimize the transmission efficiency and model accuracy.

[0102] Example 3, the above is a schematic solution of a power supply data adaptive transmission optimization method. It should be noted that the technical solution of the power supply data adaptive transmission optimization system belongs to the same concept as the technical solution of the above power supply data adaptive transmission optimization method. For the details not described in detail in the technical solution of the power supply data adaptive transmission optimization system in this embodiment, reference can be made to the description of the technical solution of the above power supply data adaptive transmission optimization method.

[0103] This embodiment also provides a power supply data adaptive transmission optimization system, including: a feature evaluation module, a data acquisition and transmission module, and an update and optimization module.

[0104] Among them, the feature evaluation module is used to obtain a dynamic data value weight matrix through data fusion and the output parameters of the business inference model, and quantitatively evaluate the value weight of features; the data acquisition and transmission module is used to adaptively adjust the sampling strategy according to the value weight of features; dynamically select the data transmission strategy according to the communication channel conditions; the update and optimization module is used to dynamically adjust the data value dynamic matrix and issue the acquisition and transmission strategy according to the change of the feature contribution value feedback by the feedback mechanism.

[0105] This embodiment also provides an electronic device, which is applicable to the situation of power supply data adaptive transmission optimization, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for realizing power supply data adaptive transmission optimization as proposed in the above embodiment.

[0106] This embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the method for realizing power supply data adaptive transmission optimization as proposed in the above embodiment.

[0107] The storage medium proposed in this embodiment and the method for realizing power supply data adaptive transmission optimization proposed in the above embodiment belong to the same inventive concept. The technical details not described in detail in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0108] Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course, it can also be implemented by hardware. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a floppy disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (FLASH), a hard disk, or an optical disc of a computer, etc., including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments of the present invention.

[0109] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. An optimized method for adaptive transmission of power supply data, characterized in that Including: Construct a data fusion and business inference model according to the power supply equipment type and the priority of the business link where it is located; Obtain a dynamic data value weight matrix through the output parameters of the data fusion and business inference model, and quantitatively evaluate the value weight of features; Adaptively adjust the sampling strategy according to the value weight of features, and dynamically select the data transmission strategy according to the communication channel conditions; Dynamically adjust the data value dynamic matrix according to the change of the feature contribution value feedback by the feedback mechanism, and issue the acquisition and transmission strategy.

2. The power supply data adaptive transmission optimization method according to claim 1, wherein The quantitative evaluation of the value weight of features includes: Normalize the event probability and business process feature data parameters output by the data fusion and business inference model; Set the business data correlation degree through the business rule library and convert it into a numerical weight coefficient; Calculate the feature timeliness coefficient based on the time sliding window, and evaluate the value of historical data through the time decay function; Perform multi-source data fusion through data analysis methods, and evaluate the model output parameters, business rules, and feature timeliness coefficients; Determine the weight allocation of each dimension, construct a dynamic data value weight matrix, and calculate the value weight of features.

3. The power supply data adaptive transmission optimization method according to claim 2, wherein The adaptive adjustment of the sampling strategy according to the value weight of features includes: Adopt a dynamic window sampling strategy for the time series features of high-value weight features, automatically adjust the sampling window according to the change gradient of features, and adjust different sampling modes according to the power supply business scenario; Adopt a sparse sampling strategy for low-value weight features.

4. The power supply data adaptive transmission optimization method according to claim 3, characterized in that, The dynamic selection of the data transmission strategy according to the communication channel conditions includes: Deploy a cyclic storage buffer on the power supply acquisition terminal side to continuously overwrite the original power supply operation data; Real-time detect the communication channel conditions, and divide the channel conditions into high, medium, and low grades according to the communication frequency band and transmission bandwidth; Adopt a hierarchical data transmission strategy according to the communication channel grade.

5. The power supply data adaptive transmission optimization method according to claim 4, wherein The change of the feature contribution value feedback by the feedback mechanism includes: Construct a data fusion model on the edge side, dynamically select the data transmission strategy, upload data on the terminal side and input it into the data fusion model; The data fusion model returns the importance of data features, dynamically adjusts the data value dynamic matrix according to the change of the feature contribution value feedback by cloud-edge collaboration, and iteratively optimizes the data feature acquisition and transmission strategy.

6. The power supply data adaptive transmission optimization method according to claim 5, wherein The sampling modes include a fault-sensitive mode, a power quality mode, and an economic operation mode.

7. The power supply data adaptive transmission optimization method according to claim 6, wherein The hierarchical data transmission strategy includes: When the terminal side is in a high-bandwidth environment, use an improved compression algorithm to transmit data losslessly; When the terminal side is in a medium-bandwidth environment, adopt a segmented compression transmission strategy; When the terminal side is in a low-bandwidth environment, adopt a feature screening strategy.

8. An optimized system for adaptive transmission of power supply data, which applies the method according to any one of claims 1-7, characterized in that Including: A feature evaluation module, a data sampling and transmission module, and an update and optimization module; The feature evaluation module is used to obtain a dynamic data value weight matrix through the output parameters of the data fusion and business inference model, and quantitatively evaluate the value weight of features; The data sampling and transmission module is used to adaptively adjust the sampling strategy according to the value weight of features; dynamically select the data transmission strategy according to the communication channel conditions; The update and optimization module is used to dynamically adjust the data value dynamic matrix according to the change of the feature contribution value feedback by the feedback mechanism, and issue the acquisition and transmission strategy.

9. An electronic device, including: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the power supply data adaptive transmission optimization method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are executed by a processor, the steps of the power supply data adaptive transmission optimization method described in any one of claims 1 to 7 are implemented.

Images