Electric energy data processing method, system and device based on DSP and multi-thread parallel computing and medium

Abstract

The invention discloses an electric energy data processing method, system and device based on a DSP and multi-thread parallel computing and a medium, and belongs to the technical field of electric energy data processing.The method comprises the steps that multi-source operation data is collected, a structured data set is constructed and written into a buffer structure, and a channel thread binding structure is generated; when the running state changes, activating the processing threads of each bound channel according to the channel thread binding structure to form a thread set; and extracting data from the current binding channel, generating a channel processing result, carrying out unified correction on a time label, constructing a state data matrix with a unified format, inputting the state data matrix into a preset evaluation model, executing operation state identification, and calculating an output result. According to the method, high-speed processing, time consistency fusion and intelligent risk identification of multi-source heterogeneous electric energy data are realized, and the method has high real-time performance and strong expansibility and is suitable for scenes such as dynamic monitoring of a power grid.

Description

Technical Field

[0001] This invention relates to the field of power data processing technology, specifically to a power data processing method, system, device, and medium based on DSP and multi-threaded parallel computing. Background Technology

[0002] As power systems evolve towards informatization and intelligence, the acquisition and processing of power data are gradually entering a stage of multi-source fusion and high-frequency sampling. Modern power grids have deployed various sensing terminals, including synchronous phasor measurement units (PMUs), SCADA systems, and fault recording equipment, to comprehensively capture key information such as power flow changes, equipment status, and transient disturbances. This data is characterized by complex structure, diverse types, and high update frequency, requiring data processing systems to possess high-speed caching, parallel processing, and cross-source alignment capabilities to support advanced functional requirements such as real-time power grid monitoring, fault early warning, and operation optimization.

[0003] However, existing power data processing solutions generally suffer from technical shortcomings such as a single processing architecture, insufficient real-time response capability, and low data fusion accuracy. On the one hand, the heterogeneity of multi-source power data in terms of time scale and data type makes traditional serial processing mechanisms unable to meet the demands of high-concurrency computing; on the other hand, existing methods lack optimization of data caching mechanisms and dynamic scheduling of processing resources, often leading to the loss of critical data or processing delays, affecting the accuracy of power grid condition assessment and risk identification. Summary of the Invention

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

[0005] Therefore, the technical problem solved by this invention is: how to construct a data cache structure with channel partitioning capability in the scenario of high-frequency generation of multi-source heterogeneous power data, and combine the interrupt triggering mechanism of DSP and multi-threaded parallel processing mode to realize structured access, channel-level parallel processing, time consistency fusion and state identification calculation of power data, thereby improving the real-time performance, accuracy and scalability of power system operation data processing.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a power data processing method based on DSP and multi-threaded parallel computing, which includes collecting multi-source operating data in the power system, classifying them uniformly according to data type and time characteristics, and constructing a structured data set with time tags;

[0007] The structured data set is written into a buffer structure with a channel partitioning mechanism, and a unique thread binding relationship is established for each data channel during the writing process to generate a channel thread binding structure.

[0008] When the running state changes, the processing threads of each bound channel are activated according to the channel thread binding structure, forming a thread set;

[0009] Each processing thread in the thread set extracts data from the currently bound channel, performs data processing operations, and generates the channel processing result.

[0010] The time stamps of the channel processing results are uniformly corrected, and the structure is fused according to the channel number and time order to construct a state data matrix in a unified format.

[0011] The state data matrix is ​​input into the preset evaluation model to perform power system operation state identification and result calculation, and the model output results containing control parameters are obtained.

[0012] As a preferred embodiment of the power data processing method based on DSP and multi-threaded parallel computing described in this invention, the step of writing the structured data set into a buffer structure with a channel partitioning mechanism includes:

[0013] When writing to a structured data set, each piece of data is mapped to a unique channel number according to the type identifier and time information recorded during the data acquisition phase, and a one-to-one binding relationship between the channel and the schedulable thread is established at the same time as the channel number is generated.

[0014] The beneficial effects of this preferred technical solution are as follows: by establishing a unique channel-thread binding relationship during the process of writing structured data sets into the buffer structure, a one-to-one mapping structure is formed, so that each data channel is handled by an independent thread in the subsequent processing, avoiding resource scheduling conflicts and thread idle problems, and effectively improving the parallel processing efficiency of the system and the logical isolation capability between data channels.

[0015] As a preferred embodiment of the power data processing method based on DSP and multi-threaded parallel computing described in this invention, wherein: the step of activating the processing threads of each bound channel according to the channel thread binding structure to form a thread set includes,

[0016] When a change in the running state is detected, the internal scheduling structure is accessed. Based on the thread binding information of each channel in the structure, the set of threads currently in the schedulable state is identified, and the thread units bound to the data channel are activated in order of scheduling priority, so that each thread establishes a real-time processing link with a unique channel.

[0017] As a preferred embodiment of the power data processing method based on DSP and multi-threaded parallel computing described in this invention, the step of each processing thread in the thread set extracting data from the currently bound channel and performing data processing operations to generate channel processing results includes:

[0018] The newly written data segments within the current time window are obtained from the buffer structure corresponding to the channel, and the data block sequence is organized according to the sampling time and channel number of each data segment. This sequence is then used as the input data structure to be processed in the internal processing flow of the thread.

[0019] As a preferred embodiment of the power data processing method based on DSP and multi-threaded parallel computing described in this invention, the construction of a unified format state data matrix includes:

[0020] Before merging the processing results of each channel, the time tag information attached to each processing result is extracted, and the time interval difference between each channel is calculated. The time axis position of the processing result is adjusted according to the difference. After aligning the time tags, the channel number and time series are arranged in a unified two-dimensional matrix structure to form a state data matrix for analysis by the preset evaluation model.

[0021] As a preferred embodiment of the power data processing method based on DSP and multi-threaded parallel computing described in this invention, the step of each processing thread in the thread set extracting data from the currently bound channel and performing data processing operations to generate channel processing results further includes:

[0022] When the processing thread performs channel data processing operations, it constructs a data time series curve according to the change of the time series sampling point interval, and calculates the channel state change trend parameters and disturbance response parameters based on the current curve;

[0023] The parameters are combined to form a feature vector, which is then used as a structure field in the channel processing results and output to the channel result set along with the channel number and corresponding time tag.

[0024] As a preferred embodiment of the power data processing method based on DSP and multi-threaded parallel computing described in this invention, the step of inputting the state data matrix into a preset evaluation model, performing power system operating state identification and result calculation, and obtaining model output results including control parameters includes,

[0025] Based on the changing trends of the time series of each channel in the state data matrix, the statistical values ​​of the phasor change rate reflecting the stability of the power grid operation are calculated and used as stability factors.

[0026] Based on the integral results of the disturbance amplitude of each channel in the state data matrix within the target analysis time window, the fluctuation energy density representing the disturbance intensity is calculated;

[0027] A weighted calculation is performed on the stability factor and the fluctuation energy density to obtain the risk level value of the operational status risk level;

[0028] The risk level value is used as the output of the assessment model.

[0029] The beneficial effects of this preferred technical solution are as follows: by performing trend identification and disturbance assessment on the state data matrix, extracting transient stability factors and fluctuation energy density, and generating risk level values ​​based on the weighted calculation of the two types of indicators, the operating state of the power system can be transformed from complex multidimensional data into risk assessment results with clear thresholds, realizing the structured, quantitative and controllable expression of state identification, providing real-time decision-making reference for subsequent control systems, and enhancing the accuracy and timeliness of operational risk response.

[0030] This invention provides an energy data processing system based on DSP and multi-threaded parallel computing.

[0031] To solve the above technical problems, the present invention provides the following technical solution: a power data processing system based on DSP and multi-threaded parallel computing, comprising: a data acquisition module, a thread binding module, a thread collection module, a channel processing module, a data matrix module, and a result output module;

[0032] The data acquisition module collects multi-source operating data from the power system and classifies it uniformly according to data type and time characteristics to construct a structured data set with time tags.

[0033] The thread binding module writes the structured data set into a buffer structure with a channel partitioning mechanism, and establishes a unique thread binding relationship for each data channel during the writing process, generating a channel thread binding structure.

[0034] The thread set module activates the processing threads of each bound channel according to the channel thread binding structure when the running state changes, forming a thread set;

[0035] The channel processing module consists of each processing thread in the thread set extracting data from the currently bound channel, performing data processing operations, and generating channel processing results.

[0036] The data matrix module performs unified correction of the time stamps on the channel processing results and completes structural fusion according to the channel number and time order to construct a state data matrix in a unified format.

[0037] The result output module inputs the state data matrix into a preset evaluation model, performs power system operation state identification and result calculation, and obtains model output results containing control parameters.

[0038] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the described method for processing electrical energy data based on DSP and multi-threaded parallel computing.

[0039] The present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the power data processing method based on DSP and multi-threaded parallel computing.

[0040] The beneficial effects of this invention are as follows: This invention introduces an interrupt triggering mechanism based on power flow switching events, and dynamically matches the number of threads and data channels, ensuring that each data channel is driven by a dedicated thread and invokes a specific processing module. This design reduces the risk of resource idleness and avoids mutual interference between algorithm modules during processing, effectively improving the resource utilization efficiency and real-time response speed of multi-threaded parallel processing, meeting the high-concurrency analysis requirements in complex power data environments. High-precision hardware time synchronization technology ensures the consistency of multi-channel processing results across the time dimension, and a trained transient stability model outputs risk indicators such as stability factors and fluctuation energy density, ultimately triggering a tiered alarm mechanism. Attached Figure Description

[0041] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 The above is a flowchart of an overall method for processing electrical energy data based on DSP and multi-threaded parallel computing, which is provided as an embodiment of the present invention. Detailed Implementation

[0043] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0044] Example 1, referring to Figure 1 As one embodiment of the present invention, this embodiment provides a power data processing method based on DSP and multi-threaded parallel computing, comprising:

[0045] S100: Collect multi-source operating data from the power system, classify them uniformly according to data type and time characteristics, and construct a structured data set with time tags.

[0046] S200. Write the structured data set into a buffer structure with a channel partitioning mechanism, and establish a unique thread binding relationship for each data channel during the writing process to generate a channel thread binding structure.

[0047] S300. When the running state changes, the processing threads of each bound channel are activated according to the channel thread binding structure to form a thread set.

[0048] S400: Each processing thread in the thread set extracts data from the currently bound channel, performs data processing operations, and generates channel processing results.

[0049] S500: Perform unified correction of time stamps on the channel processing results, and complete the structural fusion according to the channel number and time order to construct a unified format state data matrix.

[0050] S600: Input the state data matrix into the preset evaluation model, perform power system operation state identification and result calculation, and obtain the model output result containing control parameters.

[0051] It should be noted that multi-source data from PMU, SCADA system and fault recorder have characteristics such as inconsistent sampling frequency, different time series coverage, and heterogeneous data format, which can easily lead to problems such as type confusion, time misalignment and channel overlap during the raw data access stage. On the other hand, the existing data processing architecture generally adopts serial processing logic or fixed task scheduling method, which lacks a response mechanism for the dynamic concurrency characteristics of data and is difficult to adapt to the rapid processing needs under sudden conditions.

[0052] Furthermore, during data processing, multi-source channel data often suffers from buffer congestion, processing delays, or thread blocking. This is especially problematic during sudden changes in operating conditions, which can easily lead to the delay or loss of critical data, affecting the control system's real-time monitoring of the power grid's status. Even after data acquisition and processing are completed, inconsistent model input formats and fusion times further exacerbate issues such as large evaluation errors in the state recognition model, untimely responses, and unstable output control parameters.

[0053] Therefore, in response to the above-mentioned problems, this invention constructs a standardized acquisition, channel-level binding, parallel processing, time-consistent fusion, and intelligent state identification process for multi-source power data. This solves the problems of chaotic heterogeneous data access, low efficiency of processing resource scheduling, poor accuracy of channel data fusion, and insufficient state identification dimensions in traditional methods. It achieves the goal of improving the real-time performance of power system data processing, the clarity of processing paths, and the accuracy of state assessment, and is suitable for the needs of dynamic monitoring and intelligent control of power grids in complex operating scenarios.

[0054] Example 2, an embodiment of the present invention, provides a power data processing method based on DSP and multi-threaded parallel computing, based on the previous embodiment, including:

[0055] S100: Collect multi-source operating data from the power system, classify them uniformly according to data type and time characteristics, and construct a structured data set with time tags.

[0056] A preferred embodiment of constructing a time-stamped structured dataset in this invention involves: acquiring real-time voltage, current amplitude, and phase angle information from the synchronous phasor measurement unit (PMU) in the power grid to generate a PMU data subset with a global time stamp. Key operating parameters such as switch status, voltage level, and load values ​​are extracted from the dispatch control system to form a SCADA status data subset. When power disturbances or protection actions occur, high-frequency transient current and voltage curve data output from the fault recorder are acquired to generate a transient data subset. The PMU data subset, SCADA data subset, and transient data subset are each labeled with their corresponding sampling timestamps and merged to form a categorized power dataset. By categorizing and time-sequentially acquiring these three types of key data—PMU, SCADA, and fault recorder data—further processing is ensured to have a complete spatiotemporal data foundation and consistent processing characteristics.

[0057] An optional embodiment of constructing a structured data set with time tags in this invention is as follows: by accessing the unified communication gateway node already deployed in the power grid monitoring system, real-time operation data packets are collected and aggregated; the data packets contain multiple types of operation information transmitted by the front-end measurement and control unit, including analog quantity sample values, switch quantity status change information, and protection event identifiers.

[0058] After the data packet is received, based on the signal source identifier and time field contained in the data field, the values ​​of key fields such as current, voltage, and status bit are parsed and extracted, and classified into the corresponding running data subsets according to the preset data type classification rules.

[0059] The data sampling time recorded in each subset is timestamped with the central scheduling master clock and then uniformly converted into a standard format time tag.

[0060] Ultimately, all categorized data subsets will be combined to form a structured data set with a unified time base and type labels.

[0061] An optional embodiment of constructing a structured data set with time tags in this invention is as follows: by deploying local edge computing nodes on the edge side, real-time collection and preliminary processing of operational information within the substation area are performed, including locally collected analog quantities, electrical quantities, and digital alarm quantities; after receiving the raw data, the edge computing node divides the data into a subset of analog sampling data, a subset of status monitoring data, and a subset of abnormal triggering data according to the local device number and data type encoding.

[0062] Subsequently, the edge nodes generate high-precision time tags for all data subsets based on a local unified clock (such as the IEEE 1588 synchronization signal), and push these classified and time-stamped data structures to the main system through the edge bus;

[0063] After receiving data from multiple edge nodes, the main system directly integrates the data to form the final structured data set, avoiding complex parsing and processing by the central system and reducing latency and load.

[0064] It should be further explained that the beneficial effect of the preferred embodiment is that by collecting raw operating data from PMU, SCADA system and fault recorder respectively, constructing a subset of data with sampling time labels according to their respective data characteristics, and then performing time unification processing and type classification and merging, it can ensure that the constructed structured data set has high consistency on the time axis and clear source boundaries in the data dimension, avoiding the time sequence misalignment and type confusion problems caused by directly splicing heterogeneous data sources.

[0065] S200. Write the structured data set into a buffer structure with a channel partitioning mechanism, and establish a unique thread binding relationship for each data channel during the writing process to generate a channel thread binding structure.

[0066] When writing to a structured data set, each piece of data is mapped to a unique channel number according to the type identifier and time information recorded during the data acquisition phase, and a one-to-one binding relationship between the channel and the schedulable thread is established at the same time as the channel number is generated.

[0067] A preferred embodiment of the channel thread binding structure in this invention is as follows: independent data channels are allocated for different types of classified energy data, and corresponding circular buffers are established; the size of the buffer storage space is dynamically adjusted according to the sampling rate and data density of each data channel to avoid channel congestion; a write pointer and overwrite strategy is adopted to ensure that new data can overwrite old data while maintaining data integrity; a buffer overflow threshold is set, and a data export mechanism or write blocking is triggered when any channel buffer is about to be full.

[0068] By constructing a multi-channel ring buffer structure with dynamic allocation and write protection capabilities, a guarantee is provided for the stable caching and efficient retrieval of high-frequency power data.

[0069] An optional embodiment of the channel-thread binding structure in this invention is as follows: During the buffer structure initialization phase, a linear buffer of fixed capacity is uniformly set for all data channels, and the thread resources in the schedulable thread pool are polled and allocated sequentially to form a channel-thread mapping table; whenever a new category of electrical energy data is written to the corresponding channel of the buffer, the system submits the processing task to the scheduling queue based on the thread number registered in the mapping table; each thread polls the task record registered in the scheduling queue in turn, and completes data extraction and processing calls according to the channel index number; when a thread is idle, it does not actively release resources, but maintains the binding state with its corresponding channel until the buffer is cleared.

[0070] An optional embodiment of the channel thread binding structure in this invention is as follows: During the data channel initialization phase, a corresponding thread tag is preset for each type of data, and the tag is embedded into the metadata field of the structured data set; when the data is written to the buffer structure, the system parses the thread tag information in the data, and temporarily binds the current data channel to the processing thread corresponding to the tag according to the tag content;

[0071] Once a thread has finished processing its task, it releases the binding relationship and relinquishes the channel scheduling rights to the idle thread pool. The thread pool then prioritizes thread resources that match the current channel's historical binding records for rescheduling, thereby constructing a tag-driven, state-aware dynamic binding relationship.

[0072] The advantages of the preferred embodiment are: by configuring a ring buffer structure with dynamic space allocation capability for different types of power data channels, combined with the write pointer mechanism and the overlay control strategy, the system can dynamically adjust the storage capacity of each buffer according to the actual data sampling rate, thus avoiding channel congestion while ensuring the continuity of high-frequency writing; at the same time, by setting the buffer overflow threshold and the corresponding data export mechanism, the problems of buffer overflow and data truncation caused by sudden data growth are effectively avoided, thereby improving the stability and call continuity of multi-channel data caching.

[0073] S300. When the running state changes, the processing threads of each bound channel are activated according to the channel thread binding structure to form a thread set.

[0074] When a change in the running state is detected, the internal scheduling structure is accessed. Based on the thread binding information of each channel in the structure, the set of threads currently in the schedulable state is identified, and the thread units bound to the data channel are activated in order of scheduling priority, so that each thread establishes a real-time processing link with a unique channel.

[0075] Based on the PMU data change rate, SCADA switching signal and waveform recorder disturbance judgment rules, power flow switching trigger events are identified. When a power flow switching event is detected, the DSP interrupt controller starts the processing logic and sends a start signal to each processing thread channel. An equal number of parallel threads are allocated according to the number of active channels in the current ring buffer to ensure that thread scheduling is mapped one-to-one with data channels.

[0076] By combining dynamic allocation of thread resources, it enables rapid response and parallel activation of data processing resources at critical moments.

[0077] S400: Each processing thread in the thread set extracts data from the currently bound channel, performs data processing operations, and generates channel processing results.

[0078] The newly written data segments within the current time window are obtained from the buffer structure corresponding to the channel, and the data block sequence is organized according to the sampling time and channel number of each data segment. This sequence is then used as the input data structure to be processed in the internal processing flow of the thread.

[0079] When the processing thread performs channel data processing operations, it constructs a data time series curve according to the change of the time series sampling point interval, and calculates the channel state change trend parameters and disturbance response parameters based on the current curve;

[0080] The parameters are combined to form a feature vector, which is then used as a structure field in the channel processing results and output to the channel result set along with the channel number and corresponding time tag.

[0081] Each parallel processing thread is bound to a unique dedicated algorithm module to achieve a one-to-one mapping; each processing thread reads newly written data from the channel buffer it is bound to and performs data format decoding.

[0082] The corresponding algorithm module is called according to the bound data type, including phasor calculation, state discrimination and transient feature extraction, etc.; the processing results of each algorithm are re-encoded into a unified format to form a channel processing result set and cached in shared memory;

[0083] By using a thread-module binding mechanism and a differentiated algorithm processing strategy, data processing efficiency is improved, and the consistency and timeliness of feature extraction and expression for data from each channel are ensured.

[0084] S500: Perform unified correction of time stamps on the channel processing results, and complete the structural fusion according to the channel number and time order to construct a unified format state data matrix.

[0085] Before merging the processing results of each channel, the time tag information attached to each processing result is extracted, and the time interval difference between each channel is calculated. The time axis position of the processing result is adjusted according to the difference. After aligning the time tags, the channel number and time series are arranged in a unified two-dimensional matrix structure to form a state data matrix for analysis by the preset evaluation model.

[0086] It calls upon the system's built-in GPS or IEEE 1588 time synchronization unit to provide a high-precision unified clock source;

[0087] The timestamp of the channel processing result is corrected using the following formula:

[0088] T aligned =T raw +Δ sync

[0089] Among them, T raw For the original timestamp, Δ s ync represents the synchronization calibration deviation.

[0090] After aligning all processing results, they are merged according to the data source channel number and time window to generate a fusion state matrix.

[0091] By using high-precision hardware time synchronization and a unified time calibration algorithm, the time consistency and fusion accuracy of different types of data processing results are ensured.

[0092] S600: Input the state data matrix into the preset evaluation model, perform power system operation state identification and result calculation, and obtain the model output result containing control parameters.

[0093] Based on the changing trends of the time series of each channel in the state data matrix, the statistical values ​​of the phasor change rate reflecting the stability of the power grid operation are calculated and used as stability factors.

[0094] Based on the integral results of the disturbance amplitude of each channel in the state data matrix within the target analysis time window, the fluctuation energy density representing the disturbance intensity is calculated;

[0095] A weighted calculation is performed on the stability factor and the fluctuation energy density to obtain the risk level value of the operational status risk level;

[0096] The risk level value is used as the output of the assessment model.

[0097] Normalize the different physical quantities in the input fusion state matrix to adapt to the model input requirements; input the preprocessed matrix into the trained transient stability model for state identification and risk assessment.

[0098] The model output includes the stability factor S. f Wave energy density E d The risk level is quantified using the following assessment function:

[0099] R = αS f +βE d

[0100] Where α and β are the evaluation weights.

[0101] Based on the risk level thresholds set according to the assessment results, a three-level alarm signal is generated and pushed to the control system.

[0102] By integrating data-driven stability models and risk function assessments, a real-time risk quantification mechanism for power grid status is constructed to achieve proactive defensive monitoring and early warning.

[0103] Example 3 is an embodiment of the present invention. This embodiment provides an energy data processing system based on DSP and multi-threaded parallel computing, including a data acquisition module, a thread binding module, a thread collection module, a channel processing module, a data matrix module, and a result output module.

[0104] The data acquisition module collects multi-source operating data from the power system and classifies it uniformly according to data type and time characteristics to construct a structured data set with time tags.

[0105] The thread binding module writes the structured data set into a buffer structure with a channel partitioning mechanism, and establishes a unique thread binding relationship for each data channel during the writing process, generating a channel thread binding structure.

[0106] The thread set module activates the processing threads of each bound channel according to the channel thread binding structure when the running state changes, forming a thread set;

[0107] The channel processing module consists of each processing thread in the thread set extracting data from the currently bound channel, performing data processing operations, and generating channel processing results.

[0108] The data matrix module performs unified correction of the time stamps on the channel processing results and completes structural fusion according to the channel number and time order to construct a state data matrix in a unified format.

[0109] The result output module inputs the state data matrix into a preset evaluation model, performs power system operation state identification and result calculation, and obtains model output results containing control parameters.

[0110] This embodiment also provides an electronic device applicable to a power data processing method based on DSP and multi-threaded parallel computing, comprising: 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 power data processing method based on DSP and multi-threaded parallel computing proposed in the above embodiment.

[0111] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a power data processing method based on DSP and multi-threaded parallel computing as proposed in the above embodiment.

[0112] The storage medium proposed in this embodiment belongs to the same inventive concept as the method for implementing power data processing based on DSP and multi-threaded parallel computing proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0113] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, 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 computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0114] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for processing electrical energy data based on DSP and multi-threaded parallel computing, characterized in that: include, Collect multi-source operating data from the power system, classify them uniformly according to data type and time characteristics, and construct a structured data set with time tags; The structured data set is written into a buffer structure with a channel partitioning mechanism, and a unique thread binding relationship is established for each data channel during the writing process to generate a channel thread binding structure. When the running state changes, the processing threads of each bound channel are activated according to the channel thread binding structure, forming a thread set; Each processing thread in the thread set extracts data from the currently bound channel, performs data processing operations, and generates the channel processing result. The time stamps of the channel processing results are uniformly corrected, and the structure is fused according to the channel number and time order to construct a state data matrix in a unified format. The state data matrix is ​​input into the preset evaluation model to perform power system operation state identification and result calculation, and the model output results containing control parameters are obtained.

2. The power data processing method based on DSP and multi-threaded parallel computing as described in claim 1, characterized in that: The step of writing the structured data set into a buffer structure with a channel partitioning mechanism includes, When writing to a structured data set, each piece of data is mapped to a unique channel number according to the type identifier and time information recorded during the data acquisition phase, and a one-to-one binding relationship between the channel and the schedulable thread is established at the same time as the channel number is generated.

3. The power data processing method based on DSP and multi-threaded parallel computing as described in claim 2, characterized in that: The process of activating the processing threads of each bound channel according to the channel thread binding structure to form a thread set includes, When a change in the running state is detected, the internal scheduling structure is accessed. Based on the thread binding information of each channel in the structure, the set of threads currently in the schedulable state is identified, and the thread units bound to the data channel are activated in order of scheduling priority, so that each thread establishes a real-time processing link with a unique channel.

4. The power data processing method based on DSP and multi-threaded parallel computing as described in claim 3, characterized in that: The process involves each processing thread in the thread set extracting data from the currently bound channel, performing data processing operations, and generating channel processing results, including: The newly written data segments within the current time window are obtained from the buffer structure corresponding to the channel, and the data block sequence is organized according to the sampling time and channel number of each data segment. This sequence is then used as the input data structure to be processed in the internal processing flow of the thread.

5. The power data processing method based on DSP and multi-threaded parallel computing as described in claim 4, characterized in that: The construction of a unified format state data matrix includes, Before merging the processing results of each channel, the time tag information attached to each processing result is extracted, and the time interval difference between each channel is calculated. The time axis position of the processing result is adjusted according to the difference. After aligning the time tags, the channel number and time series are arranged in a unified two-dimensional matrix structure to form a state data matrix for analysis by the preset evaluation model.

6. The power data processing method based on DSP and multi-threaded parallel computing as described in claim 5, characterized in that: The process of each processing thread in the thread set extracting data from the currently bound channel, performing data processing operations, and generating the channel processing result also includes: When the processing thread performs channel data processing operations, it constructs a data time series curve according to the change of the time series sampling point interval, and calculates the channel state change trend parameters and disturbance response parameters based on the current curve; The parameters are combined to form a feature vector, which is then used as a structure field in the channel processing results and output to the channel result set along with the channel number and corresponding time label.

7. The power data processing method based on DSP and multi-threaded parallel computing as described in claim 6, characterized in that: The process involves inputting the state data matrix into a preset evaluation model, performing power system operating state identification and result calculation, and obtaining model output results including control parameters. Based on the changing trends of the time series of each channel in the state data matrix, the statistical values ​​of the phasor change rate reflecting the stability of the power grid operation are calculated and used as stability factors. Based on the integral results of the disturbance amplitude of each channel in the state data matrix within the target analysis time window, the fluctuation energy density representing the disturbance intensity is calculated; A weighted calculation is performed on the stability factor and the fluctuation energy density to obtain the risk level value of the operational status risk level; The risk level value is used as the output of the assessment model.

8. A power data processing system based on DSP and multi-threaded parallel computing, employing the power data processing method based on DSP and multi-threaded parallel computing as described in any one of claims 1 to 7, characterized in that, It includes: a data acquisition module, a thread binding module, a thread collection module, a channel processing module, a data matrix module, and a result output module; The data acquisition module collects multi-source operating data from the power system and classifies it uniformly according to data type and time characteristics to construct a structured data set with time tags. The thread binding module writes the structured data set into a buffer structure with a channel partitioning mechanism, and establishes a unique thread binding relationship for each data channel during the writing process, generating a channel thread binding structure. The thread set module activates the processing threads of each bound channel according to the channel thread binding structure when the running state changes, forming a thread set; The channel processing module consists of each processing thread in the thread set extracting data from the currently bound channel, performing data processing operations, and generating channel processing results. The data matrix module performs unified correction of the time stamps on the channel processing results and completes structural fusion according to the channel number and time order to construct a state data matrix in a unified format. The result output module inputs the state data matrix into a preset evaluation model, performs power system operation state identification and result calculation, and obtains model output results containing control parameters.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the power data processing method based on DSP and multi-threaded parallel computing as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the power data processing method based on DSP and multi-threaded parallel computing as described in any one of claims 1 to 7.

Images