Cloud edge collaboration-based power station sensor data transmission method and system

Abstract

The application discloses a power station sensor data transmission method and system based on cloud edge collaboration and belongs to the technical field of power system data communication. The application aims at the industry pain points of limited bandwidth resources of remote wind and light stations, high redundancy and large bandwidth occupation of traditional full-amount periodic transmission, adopts a lightweight mapping mechanism of measurement point identification and a measurement point sending serial number of a global self-increment integer, constructs a hybrid transmission architecture of full-amount data snapshot benchmark and differential incremental transmission, combines position implicit agreement coding and variable length integer coding to compress the message volume, and is matched with dynamic synchronization, adaptive snapshot period adjustment and a bidirectional abnormal recovery mechanism. The application significantly reduces transmission bandwidth occupation, guarantees lossless restoration and transmission reliability of time series data and adapts to the weak network operation environment of power stations.

Description

Technical Field

[0001] This invention relates to the field of power system data communication technology, and in particular to a method and system for transmitting data from power plant sensors based on cloud-edge collaboration. Background Technology

[0002] With the rapid advancement of new energy power system construction, wind farms, photovoltaic power stations, and other new energy power plants have become an important part of the power supply system. These plants are mostly deployed in remote areas and require dedicated network lines to exchange data with remote power dispatch centers. To achieve real-time monitoring and fault early warning of power equipment operation, a massive number of power equipment sensors are commonly deployed within these plants to collect multi-dimensional time-series data such as voltage, current, temperature, and power. As the scale of these plants continues to expand and the level of equipment intelligence continues to improve, the number of sensor measurement points is growing exponentially, and the data sampling frequency is also gradually increasing, leading to a sharp rise in the amount of raw data transmitted.

[0003] However, due to geographical constraints and construction costs, network bandwidth resources at remote renewable energy power plants are generally limited, and the economic costs and implementation cycles for dedicated line expansion and hardware upgrades are high. Existing technologies generally employ a full-cycle transmission mechanism, whereby the edge side reports all measurement data from all power equipment sensors to the central side at fixed intervals. This mechanism suffers from severe data redundancy; during steady-state operation of power equipment, a large number of measurement data points exhibit minimal or unchanging values, yet are still repeatedly transmitted, resulting in a significant waste of bandwidth resources.

[0004] Meanwhile, existing general data serialization and compression schemes are not customized for the slowly varying characteristics of power time-series data, resulting in limited compression efficiency. Insufficient bandwidth resources not only increase the communication operation costs of power stations but also limit the frequency of data reporting, affecting the real-time performance and reliability of core services such as equipment status monitoring and remote dispatching. Furthermore, existing cloud-edge collaborative data transmission schemes generally lack dynamic synchronization mechanisms for measurement point configuration information and bidirectional anomaly recovery mechanisms, making them prone to problems such as cloud-edge data inconsistency and data restoration failures, failing to meet the stringent requirements of power systems for data transmission accuracy and stability.

[0005] Therefore, there is an urgent need to develop an efficient, reliable, and low-overhead time-series data transmission method that is suitable for the weak network operating environment of power plants, so as to significantly reduce network transmission load and improve the utilization efficiency of existing bandwidth resources while ensuring data integrity and real-time performance. Summary of the Invention

[0006] To address the aforementioned issues, this invention proposes a cloud-edge collaborative method and system for transmitting sensor data in power plants.

[0007] To achieve the above objectives, the present invention employs the following technical solution:

[0008] A cloud-edge collaborative method for power plant sensor data transmission includes:

[0009] The edge side constructs the measurement point configuration information of all power equipment sensors in the station and sends the measurement point configuration information to the central side. The central side receives and synchronously updates the local measurement point configuration database.

[0010] The edge side collects and preprocesses the current measurement point data of each power equipment sensor according to a preset sampling period, extracts the preprocessed current measurement point data according to a preset generation period to generate a full data snapshot, and sends the full data snapshot to the center side.

[0011] Within the time window between two adjacent full data snapshots, the edge side compares the current measurement point data with the measurement point data of the previous adjacent full data snapshot. Based on the comparison result, differential processing is performed to obtain differential data. The differential data is then converted into differential messages according to a preset encoding rule and sent to the center side.

[0012] The central side decodes the differential message according to the preset encoding rules to obtain differential data, and restores it based on the latest local measurement point configuration database and full data snapshot to obtain the real acquisition time and real original value.

[0013] As a preferred embodiment of the present invention, the measurement point configuration information includes: a measurement point sending sequence number, a measurement point identifier corresponding one-to-one with the measurement point sending sequence number, and auxiliary fields; the auxiliary fields include data type, unit of measurement, Chinese name, and numerical precision.

[0014] As a preferred embodiment of the present invention, the edge side polls and detects the operating status and measurement point configuration information of all power equipment sensors in the station according to a preset clock cycle; if the addition, replacement or failure of power equipment sensors is detected, the edge side regenerates the latest measurement point configuration information of all power equipment sensors and sends it to the central side through a reliable transmission protocol, and the central side receives and synchronously updates the local measurement point configuration database.

[0015] As a preferred embodiment of the present invention

[0016] The current measurement point data includes: measurement point transmission sequence number, current acquisition time, and current acquisition value;

[0017] The preprocessing includes: converting the current acquisition time into a timestamp in a uniform format; for data whose current acquisition value is a floating-point number, converting the current acquisition value into an integer number based on the numerical precision in the auxiliary field; if the current acquisition value is an integer or string number, it is directly retained.

[0018] The full data snapshot includes the baseline acquisition value, baseline acquisition time, and measurement point transmission sequence number for each measuring point among all power equipment sensors in the station.

[0019] In a preferred embodiment of the present invention, the edge side compares the current measurement point data with the measurement point data of the previous adjacent full data snapshot, and performs differential processing based on the comparison result to obtain differential data, including:

[0020] The current acquisition time of the current measurement point data is compared with the baseline acquisition time of the same measurement point in the previous adjacent full data snapshot, and measurement points whose current acquisition time is later than the baseline acquisition time are selected.

[0021] The differential processing specifically involves: calculating the difference between the current acquisition time of the selected measurement point and the baseline acquisition time of the same measurement point sending sequence number in the previous adjacent full data snapshot, and using this as the differential acquisition time; calculating the difference between the current acquisition value of the selected measurement point and the baseline acquisition value of the same measurement point sending sequence number in the previous adjacent full data snapshot, and using this as the differential real-time value.

[0022] Generate differential data including the measurement point transmission sequence number, differential acquisition time, and differential real-time value.

[0023] As a preferred embodiment of the present invention, the step of converting the differential data into differential messages according to a preset encoding rule and sending them to the central side includes:

[0024] The encoding rules adopt positional implicit convention encoding;

[0025] The differential message includes a differential data list, which consists of one or more differential data entries; wherein the differential real-time value and differential acquisition time in the differential data are encoded using variable-length integer encoding.

[0026] As a preferred embodiment of the present invention, the central side decodes the differential message according to a preset encoding rule to obtain differential data, and restores it based on the latest local measurement point configuration database and full data snapshot to obtain the actual acquisition time and the actual original value, including:

[0027] The central side receives differential messages, performs reverse decoding using location implicit convention encoding, and extracts the measurement point transmission sequence number, differential acquisition time, and differential real-time value from the differential data; based on the latest local measurement point configuration database, it obtains the measurement point identifier and auxiliary fields that correspond one-to-one with the measurement point transmission sequence number; and based on the measurement point identifier, it determines the measurement point data to be losslessly restored in the latest full data snapshot that has been stored.

[0028] The baseline acquisition time and baseline acquisition value of the measurement point data to be losslessly restored are added to the differential acquisition time and differential real-time value in the differential data to obtain the true acquisition time and the true original value; and the data type of the true original value is restored according to the auxiliary field.

[0029] As a preferred embodiment of the present invention, if the central side detects that the full data snapshot is missing, the local measurement point configuration database is inconsistent with the measurement point configuration information of the edge side, or the lossless restoration fails, it automatically initiates a query request to the edge side, and the edge side responds and returns the latest measurement point configuration information and the full data snapshot to the central side.

[0030] As a preferred embodiment of the present invention, the edge side continuously monitors and analyzes the total number of bytes of differential packets within N transmission cycles;

[0031] If the total number of bytes is detected to be monotonically increasing over N transmission cycles, the preset generation cycle for generating a full data snapshot will be shortened by 5%.

[0032] If the total number of bytes is detected to be monotonically decreasing or fluctuating within N transmission cycles, the preset generation cycle for generating a full data snapshot will be extended by 5% or kept unchanged.

[0033] A high-efficiency data transmission system for power plant sensors based on cloud-edge collaboration includes:

[0034] A measurement point configuration information module is constructed to build the measurement point configuration information of all power equipment sensors in the station, and send the measurement point configuration information to the central side. The central side receives and updates the local measurement point configuration database synchronously.

[0035] The preprocessing and snapshot generation module collects and preprocesses the current measurement point data of each power equipment sensor according to a preset sampling period, extracts the preprocessed current measurement point data according to a preset generation period to generate a full data snapshot, and sends the full data snapshot to the central side.

[0036] The differential processing module is used to compare the current measurement point data with the measurement point data of the previous adjacent full data snapshot within the time window of two adjacent full data snapshots. Based on the comparison result, differential processing is performed to obtain differential data. The differential data is then converted into differential messages according to a preset encoding rule and sent to the central side.

[0037] The lossless restoration module is used to decode the differential message according to the preset encoding rules to obtain differential data, and restore it based on the latest local measurement point configuration database and full data snapshot to obtain the real acquisition time and real original value.

[0038] The beneficial effects of this invention are as follows: First, it significantly reduces transmission bandwidth usage. By employing a hybrid architecture of full data snapshot benchmark and differential incremental transmission, combined with implicit location coding and variable-length integer coding, bandwidth usage is significantly reduced in typical scenarios, reducing leased line communication costs and alleviating network congestion. Second, it ensures data transmission integrity and consistency. The edge side uniformly constructs the measurement point configuration information of all power equipment sensors and synchronizes it to the central side, ensuring real-time alignment of cloud-edge metadata. The central side performs lossless restoration based on the latest measurement point configuration database and full data snapshot, fully preserving the original data acquisition time and numerical accuracy. Third, it improves system real-time performance and scalability. The extreme compression of differential message volume enables higher frequency data reporting under limited bandwidth, ensuring real-time response for core services such as equipment monitoring and remote dispatching. Simultaneously, it allows access to more measurement points without increasing bandwidth investment, adapting to the needs of site expansion. Finally, it is highly adaptable to weak network environments and has strong engineering feasibility. Based on the existing cloud-edge collaborative architecture, this invention does not require large-scale upgrades to sensor hardware and network infrastructure. It can be implemented simply through software optimization, resulting in low deployment costs and short implementation cycles. It can be widely used in various wind farms and photovoltaic power stations. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the 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. Wherein:

[0040] Figure 1 This is a flowchart of the method in an embodiment of the present invention;

[0041] Figure 2 This is a system structure diagram in an embodiment of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0043] like Figure 1 As shown, this is an embodiment of the present invention, which provides a method for transmitting data from power plant sensors based on cloud-edge collaboration, including:

[0044] S1. The edge side constructs the measurement point configuration information of all power equipment sensors in the station and sends the measurement point configuration information to the central side. The central side receives and synchronously updates the local measurement point configuration database.

[0045] In this embodiment, the measurement point configuration information includes: measurement point sending sequence number, measurement point identifier corresponding one-to-one with the measurement point sending sequence number, and auxiliary fields; the auxiliary fields include data type, unit of measurement, Chinese name, and numerical precision.

[0046] The edge side polls and detects the operating status and measurement point configuration information of all power equipment sensors in the station according to a preset clock cycle. If a power equipment sensor is added, replaced, or malfunctions, the edge side regenerates the latest measurement point configuration information of all power equipment sensors and sends it to the central side through a reliable transmission protocol. The central side receives and updates its local measurement point configuration database synchronously.

[0047] Specifically, during the initialization phase, the edge side completes the acquisition and input of measurement point data for all power equipment sensors within the station. The measurement point data includes information such as the measurement point identifier, measurement point transmission sequence number, data type, unit of measurement, Chinese name, and numerical precision for each power equipment sensor. The measurement point identifier is a unique named string with hierarchical semantics, named in the format of "station code.equipment code.measurement point type", and is used by the business layer to uniquely identify and locate the power equipment sensor measurement points.

[0048] The measurement point transmission sequence number is a globally unique, monotonically increasing integer measurement point transmission sequence number for each power equipment sensor within the station. A one-to-one mapping relationship is established between the measurement point transmission sequence number and the measurement point identifier of the corresponding power equipment sensor, thus completing the lightweight mapping conversion of the measurement point identifier.

[0049] The edge side encapsulates the measurement data of each power equipment sensor's measurement point in a structured manner, forming a complete configuration entry for a single measurement point. This complete configuration entry includes the measurement point transmission sequence number corresponding to the power equipment sensor's measurement point, a measurement point identifier that corresponds one-to-one with the transmission sequence number, and auxiliary fields. The auxiliary fields include the data type, unit of measurement, Chinese name, and numerical precision of the corresponding power equipment sensor's measurement point. Data types include integer, floating-point, and string types. Numerical precision defines the number of decimal places for floating-point data, providing a benchmark for subsequent differential processing and lossless restoration of floating-point data.

[0050] The edge side collects the complete configuration entries of each single measuring point of all power equipment sensors within the station, forming the full set of measuring point configuration information for the station. The edge side performs a consistency check on the measuring point configuration information within the station. The check includes: the global uniqueness of the measuring point transmission sequence number of all power equipment sensors, the one-to-one correspondence between the mapping relationship between all measuring point transmission sequence numbers and the corresponding measuring point identifiers, and the matching of all auxiliary fields with the measuring point data attributes of the corresponding power equipment sensors. After the check passes, the edge side synchronously stores the full set of measuring point configuration information within the station into the edge side's local memory and persistent storage medium, completing the construction of the edge side's local measuring point configuration information.

[0051] After the edge side completes the construction and verification of the measurement point configuration information within the site, it initiates an initial synchronization request for the measurement point configuration information to the central side through a reliable transmission protocol, and pushes all the measurement point configuration information within the site to the central side in a preset structured message format at once.

[0052] After receiving the structured message pushed by the edge side, which includes all the configuration information of all the measurement points in the station, the central side first performs message parsing and full data verification. The verification includes: parsing the global uniqueness of the measurement point sending sequence number of all power equipment sensors, the integrity of the mapping relationship between all measurement point sending sequence numbers and corresponding measurement point identifiers, and the compliance of all auxiliary fields.

[0053] The central side completes the parsing and full data verification of the structured message containing all the measurement point configuration information within the site. After the verification is successful, the central side writes all the measurement point configuration information within the site into the local measurement point configuration database, completing the initial construction of the local measurement point configuration database. After the central side completes the construction of the local measurement point configuration database, it returns an acknowledgment message of successful initial synchronization of measurement point configuration information to the edge side through a reliable transmission protocol, completing the entire process of initial synchronization of measurement point configuration information.

[0054] The edge side performs polling detection on the operating status of all power equipment sensors and the configuration information of all measurement points in the station according to a preset clock cycle.

[0055] During the polling and detection process, if the edge side detects that a new, replaced, or faulty power equipment sensor has been added to the site, resulting in a change in the configuration information of all measurement points in the site, the edge side will immediately regenerate the latest full configuration information of all power equipment sensors in the site.

[0056] After the edge side completes the consistency verification of the latest full set of measurement point configuration information, it actively pushes the latest full set of measurement point configuration information to the central side through a reliable transmission protocol.

[0057] After receiving the latest full set of measurement point configuration information, the central side updates its local measurement point configuration database in an atomic manner, completing the synchronization and alignment of the measurement point configuration information between the edge side and the central side. After completing the update of its local measurement point configuration database, the central side returns a synchronization success confirmation message to the edge side through a reliable transmission protocol, ensuring that the measurement point configuration information between the edge side and the central side remains consistent throughout the entire operation cycle.

[0058] S2. The edge side collects and preprocesses the current measurement point data of each power equipment sensor according to the preset sampling period, extracts the preprocessed current measurement point data according to the preset generation period to generate a full data snapshot, and sends the full data snapshot to the center side.

[0059] In this embodiment, the current measurement point data includes: measurement point transmission sequence number, current acquisition time, and current acquisition value;

[0060] The preprocessing includes: converting the current acquisition time into a timestamp in a uniform format; for data whose current acquisition value is a floating-point number, converting the current acquisition value into an integer number based on the numerical precision in the auxiliary field; if the current acquisition value is an integer or string number, it is directly retained.

[0061] The full data snapshot includes the baseline acquisition value, baseline acquisition time, and measurement point transmission sequence number for each measuring point among all power equipment sensors in the station.

[0062] Specifically, after the edge side completes the initial synchronization of the measurement point configuration information of all power equipment sensors in the station, it completes the initial configuration of the full data snapshot mechanism, presets the generation cycle of the full data snapshot, core field specifications, numerical preprocessing rules, full compliance verification rules and reliable transmission rules, and clarifies that the measurement point sending sequence number must correspond one-to-one with the local measurement point configuration information.

[0063] The edge side continuously collects current measurement point data from all valid power equipment sensors within the station according to a preset sampling period, and simultaneously performs standardized preprocessing, including: replacing the measurement point identifier in the measurement point data with the corresponding globally unique integer measurement point transmission sequence number; converting the current acquisition time into a unified format millisecond-level timestamp; converting floating-point current acquisition values ​​into integer data by amplifying the numerical precision of the auxiliary field in the corresponding measurement point configuration information; if the current acquisition value is integer or string data, it is directly retained; the preprocessed measurement point data is cached locally in real time as the data source for generating the full data snapshot.

[0064] The edge side triggers a full data snapshot generation task according to a preset generation cycle, extracts the preprocessed data of all valid measurement points in the entire field at the current time, encapsulates three core fields for each measurement point: measurement point transmission sequence number, baseline acquisition value, and baseline acquisition time, and aggregates them to generate a complete full data snapshot; then, a full compliance check is performed to ensure that the measurement point transmission sequence number completely matches the local measurement point configuration information, that no core fields are missing, and that the data format is compliant; after the check passes, the full data snapshot is cached and persistently stored, and set as the baseline data source for subsequent differential processing.

[0065] The edge side encapsulates the verified full data snapshot into a dedicated transmission message through a reliable transmission protocol and pushes it to the central side. After the central side completes the integrity and compliance verification of the message, it persists the full data snapshot, updates the local snapshot cache synchronously, and sets it as the benchmark data source for subsequent lossless restoration. Finally, the central side returns a transmission success confirmation message to the edge side, completing the entire closed loop.

[0066] S3. Within the time window of two adjacent full data snapshots, the edge side compares the current measurement point data with the measurement point data of the previous adjacent full data snapshot. Based on the comparison result, differential processing is performed to obtain differential data. The differential data is converted into differential messages according to the preset encoding rules and sent to the center side.

[0067] In this embodiment, the current measurement point data includes: measurement point transmission sequence number, current acquisition time, and current acquisition value;

[0068] The current acquisition time of the current measurement point data is compared with the baseline acquisition time of the same measurement point in the previous adjacent full data snapshot, and measurement points whose current acquisition time is later than the baseline acquisition time are selected.

[0069] The differential processing specifically involves: calculating the difference between the current acquisition time of the selected measurement point and the baseline acquisition time of the same measurement point sending sequence number in the previous adjacent full data snapshot, and using this as the differential acquisition time; calculating the difference between the current acquisition value of the selected measurement point and the baseline acquisition value of the same measurement point sending sequence number in the previous adjacent full data snapshot, and using this as the differential real-time value.

[0070] Generate differential data including the measurement point transmission sequence number, differential acquisition time, and differential real-time value.

[0071] The encoding rules adopt positional implicit convention encoding;

[0072] The differential message includes a differential data list, which consists of one or more differential data entries; wherein the differential real-time value and differential acquisition time in the differential data are encoded using variable-length integer encoding.

[0073] Specifically, within the time window of two adjacent full data snapshots, the edge side continuously collects measurement point data of all valid power equipment sensors in the power station according to the preset sampling period. It performs standardized preprocessing on each measurement point data and generates current measurement point data including measurement point transmission sequence number, current acquisition time, and current acquisition value, which is cached in real time to the local memory of the power station edge side.

[0074] The power station edge side extracts the previous adjacent full data snapshot from the local cache. Using the measurement point transmission sequence number as the unique matching identifier, the current acquisition time of the current measurement point data is compared with the benchmark acquisition time corresponding to the same measurement point transmission sequence number in the previous adjacent full data snapshot. Measurement points whose current acquisition time is later than the benchmark acquisition time are selected as valid measurement points for this differential processing.

[0075] For all valid measuring points selected at the edge of the power station, differential processing is performed, including: For the time dimension, calculating the difference between the current measuring point's time and the baseline measuring time corresponding to the same measuring point's transmission sequence number in the previous adjacent full data snapshot, defining this difference as the differential measuring time; For the numerical dimension, if the current measured value is floating-point data, performing integer conversion based on numerical precision and performing numerical differential calculation, calculating the difference between the current measured value and the baseline measured value corresponding to the same measuring point's transmission sequence number in the previous adjacent full data snapshot, defining this difference as the differential real-time value; completing the calculation of differential measuring time and differential real-time value completes the differential processing for a single valid measuring point.

[0076] For all valid measuring points that have completed differential processing at the edge of the power station, structured encapsulation is performed to generate a single differential data point for each valid measuring point. The single differential data point strictly includes three core fields: measuring point transmission sequence number, differential acquisition time, and differential real-time value. All single differential data points are collected to generate a differential data list, thus completing the structured generation of the full differential data.

[0077] At the edge of the power station, the differential data list is encoded and encapsulated according to a preset encoding rule to generate differential messages. Specifically, the encoding rule adopts position implicit convention encoding. Each differential data point is encoded continuously in a fixed order of measurement point transmission sequence number, differential acquisition time, and differential real-time value. No additional field labels or type identifiers are required. The encoding and subsequent decoding are matched by the fixed position of the fields. The differential acquisition time and differential real-time value in the differential data are compressed using variable-length integer encoding to reduce the byte size of the differential results. Finally, the standardized encapsulation of the differential message is completed.

[0078] After the differential message is standardized and encapsulated at the edge of the power station, it is sent to the central side through a preset transmission protocol, thus completing the entire process of differential data processing and transmission.

[0079] S4. The central side decodes the differential message according to the preset encoding rules to obtain differential data, and restores it based on the latest local measurement point configuration database and full data snapshot to obtain the real acquisition time and real original value.

[0080] In this embodiment, the central side receives differential messages, performs reverse decoding using location implicit agreement encoding, and extracts the measurement point transmission sequence number, differential acquisition time, and differential real-time value from the differential data; based on the latest local measurement point configuration database, it obtains the measurement point identifier and auxiliary fields that correspond one-to-one with the measurement point transmission sequence number; and determines the measurement point data to be losslessly restored in the latest full data snapshot stored according to the measurement point identifier.

[0081] The baseline acquisition time and baseline acquisition value of the measurement point data to be losslessly restored are added to the differential acquisition time and differential real-time value in the differential data to obtain the true acquisition time and the true original value; and the data type of the true original value is restored according to the auxiliary field.

[0082] If the central side detects that the full data snapshot is missing, the local measurement point configuration database is inconsistent with the measurement point configuration information on the edge side, or the lossless restoration fails, it will automatically send a query request to the edge side. The edge side will respond and return the latest measurement point configuration information and the full data snapshot to the central side.

[0083] Specifically, when the central side receives differential messages sent by the edge side, it first performs a transmission integrity check on the differential messages. The check includes whether there are any issues such as data packet loss, data tampering, or abnormal message format during the transmission of the differential messages. If the transmission integrity check fails, the central side initiates a differential message retransmission request to the edge side through a reliable transmission protocol. If the integrity check passes, the central side sends the qualified differential messages into the standardized reverse decoding process.

[0084] The central side performs reverse decoding on the verified differential messages according to the preset encoding rules of this invention. Specifically, this includes: firstly, using the position implicit encoding reverse decoding rules, based on the fixed field order of the measurement point sending sequence number, differential acquisition time, and differential real-time value pre-agreed during the encoding stage, performing field-by-field parsing on the differential messages without additional parsing of field tags and type identifiers; simultaneously, performing variable-length integer reverse decoding on the differential acquisition time and differential real-time value encoded in the differential messages, and finally completely extracting the three core fields corresponding to each record in the differential data: measurement point sending sequence number, differential acquisition time, and differential real-time value, thus completing the full reverse decoding of the differential data.

[0085] The central side uses the measurement point transmission sequence number extracted by full reverse decoding as a unique index to query the latest synchronized local measurement point configuration database, obtain the measurement point identifier that is globally unique and strictly one-to-one with the measurement point transmission sequence number, as well as the auxiliary fields of the corresponding measurement point. The auxiliary fields include data type, unit of measurement, Chinese name, and numerical precision. At the same time, the central side uses the obtained measurement point identifier as the unique matching basis to locate the measurement point data to be restored without loss in the latest full data snapshot that has been persistently stored locally. The measurement point data includes the benchmark acquisition time and benchmark acquisition value of the corresponding measurement point.

[0086] The central side performs a lossless restoration calculation based on the positioning completed measurement point data and the corresponding differential data obtained from reverse decoding. Specifically, this includes: for the time dimension, summing the baseline acquisition time of the measurement point data to be losslessly restored with the differential acquisition time in the differential data with the same measurement point transmission sequence number to obtain the actual acquisition time of the corresponding measurement point; for the numerical dimension, summing the baseline acquisition value of the measurement point data to be losslessly restored with the differential real-time value in the corresponding differential data with the same measurement point transmission sequence number to obtain the actual original value of the corresponding measurement point.

[0087] Based on the auxiliary fields of the corresponding measurement points obtained from the latest local measurement point configuration database, the central side performs data type compliance restoration on the restored true original values. This includes: if the baseline acquisition value of the corresponding measurement point is an integer converted from floating-point data after numerical precision amplification, the central side divides the true original value obtained from the basic restoration by the amplification factor corresponding to the numerical precision, according to the numerical precision in the auxiliary field of the corresponding measurement point, to restore the integer value to a floating-point true original value with the same precision as the original acquisition value; if the baseline acquisition value of the corresponding measurement point is a raw integer or string data, the central side directly retains the true original value obtained from the basic restoration without additional precision conversion.

[0088] The central side performs a lossless restoration compliance check on the actual acquisition time and original values ​​of the completed full-process restoration. The check includes the consistency of data type and measurement point configuration information, the matching of numerical accuracy and auxiliary fields, and the continuity of timing logic. After the check passes, the central side associates the restored original values ​​with the corresponding measurement point identifiers and auxiliary fields and archives them for persistent storage and subsequent business processing. If a lossless restoration failure is detected during the check, the central side immediately triggers the pre-set bidirectional anomaly recovery mechanism of this invention, and sends a query request to the edge side for the latest measurement point configuration information and full data snapshot. After data synchronization is completed, the lossless restoration operation is re-executed to ensure the integrity and accuracy of the data restoration.

[0089] In this embodiment, the edge continuously monitors and analyzes the total number of bytes of differential packets within N transmission cycles;

[0090] If the total number of bytes is detected to be monotonically increasing over N transmission cycles, the preset generation cycle for generating a full data snapshot will be shortened by 5%.

[0091] If the total number of bytes is detected to be monotonically decreasing or fluctuating within N transmission cycles, the preset generation cycle for generating a full data snapshot will be extended by 5% or kept unchanged.

[0092] Specifically, after the power station edge side completes the generation of a full data snapshot and the initial operation of the differential transmission mechanism, it completes the parameter configuration of the adaptive adjustment mechanism, presets the number of transmission cycles N for continuous monitoring, the threshold for judging the fluctuation amplitude of the total number of bytes of differential messages, the initial generation cycle of the full data snapshot, the upper and lower limit thresholds for cycle adjustment, and the statistical rules for the total number of bytes of differential messages in a single transmission cycle; in this embodiment, N is 10.

[0093] At the end of each differential message transmission cycle, the edge side of the power station counts the total number of bytes of all differential messages sent to the central side during that cycle, generates single-cycle statistical data, and archives and stores it in conjunction with cycle information; at the same time, it retains only the latest N consecutive transmission cycles of valid statistical data as the data source for trend analysis.

[0094] The edge side of the power station extracts the total number of differential messages in the latest N consecutive transmission cycles and performs trend determination according to preset rules, including: if the total number of differential messages in the dataset is monotonically increasing in the N consecutive transmission cycles, a shortening operation of the full data snapshot generation cycle is triggered; if the total number of differential messages in the dataset is monotonically decreasing, a extending operation of the full data snapshot generation cycle is triggered; if the fluctuation range of the dataset does not exceed a preset threshold, a cycle maintenance operation is triggered.

[0095] Based on trend determination results, the edge side of the power station performs standardized periodic adjustments according to the rules of this invention. Specifically, this includes: in a monotonically increasing state, shortening the preset period for generating full data snapshots by 5%; in a monotonically decreasing state, extending the preset period for generating full data snapshots by 5%; and maintaining the original preset period unchanged in a stable fluctuation state. During the adjustment process, the synchronous verification period does not exceed the preset upper and lower thresholds, and the adjusted period officially takes effect at the next full data snapshot generation node. The edge side continuously executes the entire process of statistics, analysis, and adjustment, forming a closed-loop adaptive optimization to always maintain the optimal state of transmission bandwidth utilization.

[0096] The customized differential encoding transmission scheme based on variable-length integer encoding adopted in this invention is compared with the native serialization encoding scheme of Protocol Buffers (hereinafter referred to as protobuf encoding scheme) commonly used in the field of industrial IoT through mathematical modeling and formula derivation. The core comparison dimensions are the byte occupancy of single measurement point transmission and the total bandwidth occupancy of full measurement point transmission in a single cycle. The transmission efficiency improvement effect of the scheme of this invention is quantitatively verified, specifically including:

[0097] 1. The core application scenarios of this invention are wind power plants and photovoltaic power stations in the power industry (hereinafter referred to as power wind and solar power stations). The number of power equipment sensor collection points deployed in a single station of this type of station is usually in the range of 10,000 to 100,000, which fully matches the technical adaptation range of this invention. Therefore, this model sets the total number of effective collection points in a single station to n, and satisfies 10,000 < n < 100,000, covering all typical application scenarios of this invention.

[0098] 2. The power industry commonly uses a measurement point identifier format: "Station Code.Equipment Code.Measurement Point Type". The hierarchical string naming convention for measurement point identifiers is illustrated in the example: fdz01.device05.temperature. This identifier has business semantic readability, but its string length is long and its transmission byte usage is high. This model sets the string length of a single measurement point identifier as variable 'a'. One of the core innovative designs of this invention is to replace the long string of measurement point identifiers with globally incrementing integer measurement point transmission sequence numbers for data transmission, thereby fundamentally reducing the transmission byte overhead of the measurement point identifier dimension.

[0099] 3. The data collected by sensors on power equipment in power, wind, and solar power plants cover mainstream numerical types such as variable-length integers, single-precision floating-point numbers, and double-precision floating-point numbers. The serialized value of a single collected value occupies 1 to 8 bytes. This model sets the probability that a single collected value occupies i bytes after serialization as Pᵢ (i=1,2,…,8). The sum of probabilities for all data types satisfies normalization constraints. Furthermore, this invention designs an integer conversion preprocessing rule based on numerical precision for floating-point data. This rule converts floating-point data to integers and performs differential encoding, further adapting to the compression characteristics of variable-length integer encoding and providing a foundation for bandwidth optimization.

[0100] 4. Calculation rules for the number of bytes transmitted under two encoding schemes in a single acquisition cycle and derivation of the formula for the total number of bytes.

[0101] 4.1 The calculation rules for the number of bytes transmitted under the two encoding schemes are shown in Table 1:

[0102] Table 1. Rules for calculating the number of bytes transmitted under two encoding schemes.

[0103]

[0104] 4.2 Derivation of the formula for the total number of bytes transmitted in a single acquisition cycle

[0105] Based on the above itemized calculation rules, the formulas for calculating the total number of bytes transmitted across all measurement points in a single acquisition cycle for both encoding schemes are derived, including:

[0106] (1) Total number of bytes in the protobuf encoding scheme The formula is:

[0107]

[0108] In the formula, This represents the total number of valid power equipment sensor measurement points within the power station. The character length of the hierarchical string for measuring point identification in the power industry standard. The serialized values ​​collected at a single measurement point occupy [a certain amount of space]. The probability of a byte. The length of bytes occupied after serializing the collected value.

[0109] (2) Total number of bytes in the differential encoding scheme of the present invention The formula is:

[0110]

[0111] Based on the above two encoding schemes, the formulas for calculating the total number of bytes transmitted across all measurement points in a single acquisition cycle are expanded and simplified. The ratio of the total number of bytes transmitted (i.e., bandwidth utilization) between the differential encoding scheme and the protobuf encoding scheme of this invention is then calculated, yielding the following quantization formula:

[0112]

[0113] In the formula, The character length for the hierarchical string identifier of the power industry standard measurement point adopts the standardized naming rule of "station code.equipment code.measurement point type", for example: fdz01.device05.temperature. To ensure business semantic recognizability, it usually includes hierarchical redundant information, and the total character length of a single measurement point identifier is usually not less than 17 characters. This model takes a conservative and general approach to ensure the quantification results. The conservative assumption is 14. Substituting 14 into the above formula, we obtain the simplified formula for bandwidth utilization ratio quantization:

[0114]

[0115] Based on the above formula, a limit boundary analysis is performed on the bandwidth optimization effect of the present invention:

[0116] Denominator extreme value analysis: When the probability distribution of the occupancy of the collected value bytes satisfies P8=1 and the rest Pi=0 (i=1,2,…,7), the denominator of the formula reaches the maximum value of 38; when the probability distribution of the occupancy of the collected value bytes satisfies P1=1 and the rest Pi=0 (i=2,3,…,8), the denominator of the formula reaches the minimum value of 31.

[0117] Molecular extreme value analysis: In this modeling, the range of the differential real-time value transmission byte number coefficient x and the differential acquisition time transmission byte number coefficient y is [1,3]. Therefore, the maximum value of x+y in the molecule is 6 and the minimum value is 2.

[0118] Based on the above extreme value analysis, the bandwidth occupancy range of the differential coding scheme of the present invention relative to the protobuf coding scheme can be clearly identified. The transmission efficiency improvement of the present invention can be derived in reverse, and the bandwidth optimization effect of the present invention can be verified, as shown in Table 2:

[0119] Table 2. Comparison of transmission performance between the differential coding scheme and the protobuf coding scheme of this invention.

[0120]

[0121] Based on the above analysis of extreme boundary conditions across all scenarios, the differential coding scheme of this invention achieves a transmission efficiency improvement of at least 70% compared to the protobuf coding scheme, significantly reducing bandwidth consumption for power plant data transmission. Furthermore, the bandwidth optimization effect of this differential coding scheme exhibits a positive amplification characteristic: the longer the string length of the hierarchical measurement point identifier in the power plant, the more significant the byte compression benefit of replacing the long string measurement point identifier with a globally incrementing integer measurement point transmission sequence number, resulting in a higher overall transmission efficiency improvement.

[0122] To intuitively verify the bandwidth compression performance of the transmission scheme proposed in this invention, temperature measurement points of typical equipment in power, wind, and solar power plants were selected to conduct a comparative verification of the coding effect at single measurement points. The verification process strictly followed the technical rules of this invention, as detailed below:

[0123] The measurement point data and protobuf encoding scheme results include:

[0124] In this verification, the hierarchical string identifier for the measurement point was selected as cz1.device1.temperature. The baseline acquisition value corresponding to the previous adjacent full data snapshot of this measurement point was 22.1℃, and the baseline acquisition time was 1772428800495 (millisecond-level Unix timestamp). The measurement point data of this point was encoded using the protobuf encoding scheme. The encoding result is: 0A17637A312E646576696365312E74656D706572617475726515CDCCA84118EFE3F4E7CA33, with a total of 37 bytes after encoding.

[0125] The differential coding scheme processing flow and coding results of this invention include:

[0126] The operation is performed following the differential processing and encoding rules. The specific steps are as follows:

[0127] First, based on the measurement point configuration information on the edge side, the global auto-incrementing integer measurement point sending sequence number corresponding to this measurement point is determined to be 2000, and the numerical precision in the auxiliary field of this measurement point is 1 decimal place;

[0128] Secondly, the current measured value at this point is 22.2℃, and the current measurement time is 1772428801495 (millisecond-level Unix timestamp). According to the floating-point data preprocessing rules of this invention, the reference measured value and the current measured value are converted into integer data by multiplying the numerical precision by 10, and then differential calculation is performed: the real-time differential value is 1 (corresponding to the original numerical difference of 0.1℃), and the differential measurement time is 1000 milliseconds (the difference between the current measurement time and the reference measurement time).

[0129] The third step involves using the implicit positional encoding and variable-length integer encoding defined in this invention to encapsulate and encode the measurement point transmission sequence number, differential acquisition time, and differential real-time value in a fixed order. The final encoding result is D00F01E807, and the total number of bytes after encoding a single measurement point is 5 bytes.

[0130] Comparative verification shows that, for typical steady-state operation measurement points of power, wind, and solar power plants, the differential coding scheme of this invention reduces the number of bytes transmitted per measurement point from 37 bytes to 5 bytes compared to the protobuf coding scheme, reducing bandwidth usage by 86.5%. This directly verifies the extreme bandwidth compression effect of the scheme of this invention, which is completely consistent with the performance improvement conclusions derived from the mathematical modeling above. It can effectively solve the industry pain points of limited bandwidth resources and high redundancy of traditional transmission methods in power, wind, and solar power plants.

[0131] like Figure 2 As shown, this is the second embodiment of the present invention. This embodiment provides a high-efficiency transmission system for power plant sensor data based on cloud-edge collaboration, including:

[0132] A measurement point configuration information module is constructed to build the measurement point configuration information of all power equipment sensors in the station, and send the measurement point configuration information to the central side. The central side receives and updates the local measurement point configuration database synchronously.

[0133] The preprocessing and snapshot generation module collects and preprocesses the current measurement point data of each power equipment sensor according to a preset sampling period, extracts the preprocessed current measurement point data according to a preset generation period to generate a full data snapshot, and sends the full data snapshot to the central side.

[0134] The differential processing module is used to compare the current measurement point data with the measurement point data of the previous adjacent full data snapshot within the time window of two adjacent full data snapshots. Based on the comparison result, differential processing is performed to obtain differential data. The differential data is then converted into differential messages according to a preset encoding rule and sent to the central side.

[0135] The lossless restoration module is used to decode the differential message according to the preset encoding rules to obtain differential data, and restore it based on the latest local measurement point configuration database and full data snapshot to obtain the real acquisition time and real original value.

[0136] In summary, this invention significantly reduces transmission bandwidth usage. Through a lightweight mapping of globally incrementing integer measurement point transmission sequence numbers and measurement point identifiers, a hybrid architecture of full data snapshot benchmark and differential incremental transmission, and collaborative optimization of implicit location encoding and variable-length integer encoding, it fully adapts to the slowly changing characteristics of sequence data during steady-state operation of power equipment. In typical scenarios, transmission bandwidth usage is reduced by more than 70%, significantly reducing dedicated line communication costs and alleviating network congestion in weak network environments. Secondly, it improves real-time performance and scalability. Through extreme compression of differential message volume, it can support more measurement point access with limited bandwidth resources, meeting the equipment expansion needs of power wind and solar power plants. Simultaneously, it can increase data reporting frequency, ensuring the real-time performance and stability of core services such as equipment monitoring and remote dispatching. Finally, it ensures data transmission consistency and reliability. Through dynamic synchronization of measurement point configuration information and a two-way anomaly recovery mechanism, it achieves real-time alignment of data between the edge and center sides. Through an adaptive full data snapshot cycle adjustment mechanism, it dynamically adapts to data fluctuation characteristics, avoiding differential cumulative errors and ensuring the system always maintains optimal transmission efficiency. It can be implemented without hardware upgrades and has strong engineering practicality.

[0137] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any other combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product, which includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0138] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. This storage medium can be a read-only memory, a disk, or an optical disk, etc.

[0139] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A power plant sensor data transmission method based on cloud-edge collaboration, characterized in that, include: The edge side constructs the measurement point configuration information of all power equipment sensors in the station and sends the measurement point configuration information to the central side. The central side receives and synchronously updates the local measurement point configuration database. The edge side collects and preprocesses the current measurement point data of each power equipment sensor according to the preset sampling period, and extracts the preprocessed current measurement point data to generate a full data snapshot according to the preset generation period. Send the full data snapshot to the central side; Within the time window between two adjacent full data snapshots, the edge side compares the current measurement point data with the measurement point data of the previous adjacent full data snapshot. Based on the comparison result, differential processing is performed to obtain differential data. The differential data is then converted into differential messages according to a preset encoding rule and sent to the center side. The central side decodes the differential message according to the preset encoding rules to obtain differential data, and restores it based on the latest local measurement point configuration database and full data snapshot to obtain the real acquisition time and real original value.

2. The power station sensor data transmission method based on cloud-edge collaboration according to claim 1, characterized in that, The measurement point configuration information includes: measurement point sending sequence number, measurement point identifier corresponding one-to-one with the measurement point sending sequence number, and auxiliary fields; the auxiliary fields include data type, unit of measurement, Chinese name, and numerical precision.

3. The power station sensor data transmission method based on cloud-edge collaboration according to claim 1, characterized in that, The edge side polls and detects the operating status and measurement point configuration information of all power equipment sensors in the station according to a preset clock cycle; If a power equipment sensor is added, replaced, or malfunctions, the edge side regenerates the latest measurement point configuration information for all power equipment sensors and sends it to the central side via a reliable transmission protocol. The central side receives the information and synchronously updates its local measurement point configuration database.

4. The power station sensor data transmission method based on cloud-edge collaboration according to claim 2, characterized in that, The current measurement point data includes: measurement point transmission sequence number, current acquisition time, and current acquisition value; The preprocessing includes: converting the current acquisition time into a timestamp in a uniform format; for data whose current acquisition value is a floating-point number, converting the current acquisition value into an integer number based on the numerical precision in the auxiliary field; if the current acquisition value is an integer or string number, it is directly retained. The full data snapshot includes the baseline acquisition value, baseline acquisition time, and measurement point transmission sequence number for each measuring point among all power equipment sensors in the station.

5. The power station sensor data transmission method based on cloud-edge collaboration according to claim 4, characterized in that, The edge side compares the current measurement point data with the measurement point data of the previous adjacent full data snapshot. Based on the comparison result, it performs differential processing to obtain differential data, including: The current acquisition time of the current measurement point data is compared with the baseline acquisition time of the same measurement point in the previous adjacent full data snapshot, and measurement points whose current acquisition time is later than the baseline acquisition time are selected. The differential processing specifically involves: calculating the difference between the current acquisition time of the selected measurement point and the baseline acquisition time of the same measurement point sending sequence number in the previous adjacent full data snapshot, and using this as the differential acquisition time; calculating the difference between the current acquisition value of the selected measurement point and the baseline acquisition value of the same measurement point sending sequence number in the previous adjacent full data snapshot, and using this as the differential real-time value. Generate differential data including the measurement point transmission sequence number, differential acquisition time, and differential real-time value.

6. The power station sensor data transmission method based on cloud-edge collaboration according to claim 5, characterized in that, The step of converting the differential data into differential messages according to a preset encoding rule and sending them to the central side includes: The encoding rules adopt positional implicit convention encoding; The differential message includes a differential data list, which consists of one or more differential data entries; wherein the differential real-time value and differential acquisition time in the differential data are encoded using variable-length integer encoding.

7. The power station sensor data transmission method based on cloud-edge collaboration according to claim 6, characterized in that, The central side decodes the differential message according to a preset encoding rule to obtain differential data, and restores it based on the latest local measurement point configuration database and full data snapshot to obtain the actual acquisition time and the actual original values, including: The central side receives differential messages, performs reverse decoding using implicit location encoding, and extracts the measurement point transmission sequence number, differential acquisition time, and differential real-time value from the differential data; based on the latest local measurement point configuration database, it obtains the measurement point identifier and auxiliary fields that correspond one-to-one with the measurement point transmission sequence number; and determines the measurement point data to be losslessly restored in the latest full data snapshot stored according to the measurement point identifier. The baseline acquisition time and baseline acquisition value of the measurement point data to be losslessly restored are added to the differential acquisition time and differential real-time value in the differential data to obtain the true acquisition time and the true original value; and the data type of the true original value is restored according to the auxiliary field.

8. The power station sensor data transmission method based on cloud-edge collaboration according to claim 1, characterized in that, If the central side detects that the full data snapshot is missing, the local measurement point configuration database is inconsistent with the measurement point configuration information on the edge side, or the lossless restoration fails, it will automatically send a query request to the edge side. The edge side will respond and return the latest measurement point configuration information and the full data snapshot to the central side.

9. The power station sensor data transmission method based on cloud-edge collaboration according to claim 1, characterized in that, edge The system continuously monitors and analyzes the total number of bytes in differential packets over N transmission cycles. If the total number of bytes is detected to be monotonically increasing over N transmission cycles, the preset generation cycle for generating a full data snapshot will be shortened by 5%. If the total number of bytes is detected to be monotonically decreasing or fluctuating within N transmission cycles, the preset generation cycle for generating a full data snapshot will be extended by 5% or kept unchanged.

10. A high-efficiency data transmission system for power plant sensors based on cloud-edge collaboration, employing the data transmission method for power plant sensors based on cloud-edge collaboration as described in claims 1-9, characterized in that... include: A measurement point configuration information module is constructed to build the measurement point configuration information of all power equipment sensors in the station, and send the measurement point configuration information to the central side. The central side receives and updates the local measurement point configuration database synchronously. The preprocessing and snapshot generation module collects and preprocesses the current measurement point data of each power equipment sensor according to the preset sampling period, and extracts the preprocessed current measurement point data to generate a full data snapshot according to the preset generation period. Send the full data snapshot to the central side; The differential processing module is used to compare the current measurement point data with the measurement point data of the previous adjacent full data snapshot within the time window of two adjacent full data snapshots. Based on the comparison result, differential processing is performed to obtain differential data. The differential data is then converted into differential messages according to a preset encoding rule and sent to the central side. The lossless restoration module is used to decode the differential message according to the preset encoding rules to obtain differential data, and restore it based on the latest local measurement point configuration database and full data snapshot to obtain the real acquisition time and real original value.

Images