Real-time acquisition and storage method for key operating parameters of frequency converter

Abstract

The invention discloses a real-time acquisition and storage method for key operating parameters of frequency converters, which belongs to the technical field of frequency converters, and comprises the following steps: aiming at the characteristics of concurrent access of multiple frequency converters, various data types, out-of-order arrival and continuous growth, performing acquisition, time alignment and packaging according to the data types on the operating parameters at an edge side; and a logic storage address is constructed on the center side based on the time interval, the equipment identifier and the data type, and mapping and ordered writing of the logic address and a physical storage position are realized through a storage positioning mechanism, so that the problem that the operation data of the plurality of frequency converters cannot be uploaded under the conditions of high concurrency and disordered uploading is solved. The technical problems of complete acquisition, stable storage and rapid positioning query are solved, the integrity of the original operation data is ensured, the network transmission pressure is effectively reduced, storage disorder caused by an uncertain data arrival sequence is avoided, and the storage efficiency and query efficiency of the operation data of the large-scale frequency converter are improved.

Description

Technical Field

[0001] This invention belongs to the field of frequency converter technology, and in particular relates to a method for real-time acquisition and storage of key operating parameters of frequency converters. Background Technology

[0002] As a key actuator in industrial automation systems, the frequency converter's critical operating parameters, such as current, voltage, and frequency, directly reflect its operating status and quality. Long-term, continuous acquisition, storage, and analysis of these critical operating parameters are essential for equipment operation monitoring, fault diagnosis, and operational optimization.

[0003] In the existing technology, there are several main implementation methods for the acquisition and storage of inverter operation data: one type of solution adopts a centralized acquisition and storage structure, in which the monitoring center directly acquires and stores the operation data of multiple inverters in real time; another type of solution adopts the method of preprocessing or filtering at the edge side before uploading, and only uploads a portion of the processed data to the central side for storage and analysis.

[0004] However, the aforementioned existing technical solutions still have significant shortcomings in practical applications: On the one hand, centralized data acquisition is prone to data transmission instability due to limited network bandwidth and data write conflicts when facing scenarios with a large number of frequency converters operating simultaneously, making it difficult to support the demand for high concurrency and long-term continuous data acquisition. On the other hand, while edge-side preprocessing or filtering upload can reduce the data pressure on the central side, it often results in incomplete raw operating data, affecting subsequent refined analysis and historical traceability.

[0005] In addition, existing technologies generally rely on the order of data arrival or timestamps to directly write and query operating data. When data from different frequency converters arrives at the monitoring center in an out-of-order and high-concurrency manner, problems such as data order disorder, difficulty in storage and location, and decreased query efficiency are likely to occur, making it difficult to achieve efficient management and rapid retrieval of large-scale frequency converter operating data. Summary of the Invention

[0006] The purpose of this invention is to provide a method for real-time acquisition and storage of key operating parameters of frequency converters, which solves the technical problem of achieving complete acquisition, stable storage and rapid location query under the conditions of high concurrency and out-of-order uploading of operating data from multiple frequency converters.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: A method for real-time acquisition and storage of key operating parameters of a frequency converter includes the following steps: Step 1: The acquisition unit set up on the inverter side collects the operating data in real time during the operation of the inverter. Each acquisition unit contains an edge storage unit. The acquisition unit sorts the operating data by time according to the dimension of a single inverter and then stores it in the edge storage unit. The acquisition unit divides the operating data in the edge storage unit into blocks according to time intervals, generates multiple data blocks, and adds source information to each data block. Step 2: When the preset triggering conditions are met, the acquisition unit selects at least one complete data block from the edge storage unit as the object to be uploaded, and splits the running data contained in the selected data block according to the data type. The running data of the same data type in the same data block are reassembled into the corresponding edge data packets. The acquisition unit uploads the edge data packets to the monitoring center asynchronously through the network. Step 3: The monitoring center receives and parses the received edge-side data packets, extracts the metadata associated with the running data, generates logical storage addresses based on the metadata, establishes a storage location mapping between the logical storage addresses and data metadata, and generates a data logical location structure independent of the physical storage location. Step 4: The monitoring center writes the corresponding running data into the database server according to the logical storage address and storage location mapping. When performing the write operation, it supports out-of-order writing and concurrent writing, and updates the storage status information after the writing is completed. Step 5: After the client sends a data query request to the monitoring center, the monitoring center filters and locates the logical storage address according to the storage location mapping, and reads the corresponding operating data from the database server according to the filtering and location results, thus completing the fast query of the operating data of the specified frequency converter.

[0008] Preferably, each acquisition unit is an edge node, and each edge node is responsible for acquiring the operating data of one frequency converter. The types of operating data include current, voltage and frequency. The acquisition unit consists of an internal sampling module of the frequency converter or a sensor module connected to it, and a data processing module; the edge-side storage unit is the local memory of the acquisition unit.

[0009] Preferably, step 1 includes the following steps: Step 1-1: During the operation of the frequency converter, the edge nodes continuously collect the operating data of the frequency converter and uniformly time-stamp the collected operating data to form a raw data sequence with time consistency. i (t k Specifically, it is expressed as: Data i (t k )={I(t k),U(t k ),f(t k ),…}; Among them, t k Indicates the time; I, U, and f represent current, voltage, and frequency, respectively; i represents the inverter number. Step 1-2: Convert the original data sequence Data i (t k When writing to edge-side storage cells, the write order must satisfy the following conditions: Data i (t1)→Data i (t2)→⋯→Data i (t n ), t1<t2<...<t n ; Where n represents the last moment; Steps 1-3: According to a preset fixed time interval, the continuously written running data is segmented to generate multiple consecutive data blocks; Steps 1-4: Add corresponding metadata to each data block. i,m Specifically, it is expressed as follows: Meta i,m ={DeviceID,DataTypeSet,T start ,T end ,SeqNo}; Where DeviceID represents the inverter identifier; DataTypeSet represents the collection of all types of operating data contained in this data block; T start Indicates the start time of data collection; T end This indicates the end time of data collection; SeqNo indicates the sequence number of the locally generated sequence.

[0010] Preferably, step 2 includes the following steps: Step 2-1: When the preset triggering conditions are met, the edge node selects one or more complete data blocks from the edge-side storage unit as the objects to be uploaded; The preset trigger conditions are: triggering at a fixed time interval, triggering when the cumulative data volume reaches a threshold, or triggering when the local cache capacity reaches a threshold. Step 2-2: For each selected data block, split the running data contained therein according to the type of running data; extract the running data of the same data type in the same data block to form the corresponding data type subsequence; Steps 2-3: For each data type subsequence, combine it with the metadata of the original data block to encapsulate and generate the corresponding edge-side data packet. The edge-side data packet contains the running data sequence of a single data type, the inverter identifier DeviceID, the data type DataType, and the data acquisition start time T. start Data collection end time T end and the local sequence number SeqNo; Steps 2-4: The acquisition unit will asynchronously upload multiple edge-side data packets to the monitoring center via the network. Edge-side data packets from different frequency converters, different data types, or different time intervals will arrive at the monitoring center in an out-of-order or concurrent manner.

[0011] Preferably, step 3 includes the following steps: Step 3-1: The monitoring center receives edge-side data packets uploaded from multiple edge nodes, and parses each edge-side data packet. Specifically, it separates the running data from the metadata describing the source and time of the running data, and marks the frequency converter, acquisition time range and data type corresponding to the edge-side data packet. Step 3-2: Based on the data collection start time contained in the edge-side data packets, the monitoring center maps the edge-side data packets to a preset time interval, so that the data on the continuous time axis is logically divided into multiple time periods. Specifically, a time interval length ΔT is preset, and the continuous time axis is discretized into time buckets of fixed length. The time buckets are represented as follows: TimeBucket=⌊T star / ΔT⌋; Among them, T star Indicates the start time of data collection, specifically obtained from metadata; TimeBucket is the time bucket. Step 3-3: Within the same time interval, for edge-side data packets originating from the same frequency converter and belonging to the same data type, the monitoring center assigns a logical segment identifier to the edge-side data packet to describe the logical sequence relationship of the edge-side data packet among similar data, specifically as follows: SegmentID=SeqID or AutoIncrement(TimeBucket, DeviceID, DataType) Where DeviceID represents the inverter identifier, DataType represents the data type, SeqID represents the sequence number of the edge node, SegmentID represents the logical segment identifier, and AutoIncrement() represents the automatic increment function, which is a sequential numbering rule for assigning incrementing logical segment identifiers to different data segments under the same time interval, device identifier, and data type conditions. DeviceID, DataType, and SeqID are all obtained from metadata; Steps 3-4: The monitoring center concatenates the time bucket, inverter identifier, data type, and logical segment identifier to generate a unique logical storage address. This logical storage address is used to describe only the logical location of the running data in the edge-side data packet under the [time + device + type] dimension. The logical storage address is represented as follows: LogicalAddress=<TimeBucket,DeviceID,DataType,SegmentID> ; Wherein, LogicalAddress is the logical storage address; Steps 3-5: The monitoring center associates the generated logical storage addresses with the corresponding runtime data metadata and writes them into a storage location mapping, which serves as the basic index information for data writing and querying. The storage location mapping is represented as follows: MappingRecord={LogicalAddress, DeviceID, DataType, T start T end ,Fs,N} Where MappingRecord is the storage location mapping, T end T is the end time of data acquisition, Fs is the sampling frequency, and N is the number of data points; end Both Fs and N are obtained from the metadata; Steps 3-6: The monitoring center establishes a storage location mapping list based on the storage location mapping, and constructs a data logical location structure independent of the physical storage location; The physical storage location refers to the physical storage address location of the database server.

[0012] Preferably, step 4 includes the following steps: Step 4-1: The monitoring center uses the logical storage address and storage location mapping as the sole basis for data writing, triggering the data writing process. Specifically, based on the time interval identifier, device identifier, and data type in the logical storage address, the running data to be written is mapped to the corresponding storage partition in the database server, thus associating the logical storage structure with the physical storage structure to determine the physical storage location. This is illustrated as follows: PhysicalLocation=f(TimeBucket, DeviceID, DataType); Wherein, PhysicalLocation represents the association between logical storage structure and physical storage structure; TimeBucket is the time bucket; DeviceID represents the inverter identifier; DataType represents the data type; f() represents the mapping function, which is a mapping rule for selecting the physical storage location of running data based on the time interval identifier, device identifier and data type in the logical storage address; Step 4-2: After determining the physical storage location, organize the running data under the same logical storage address sequentially according to the logical segment identifier (SegmentID), and write the running data into the corresponding storage space in sequence. The sequential organization is represented as follows: WriteOrder ∝ SegmentID; WriteOrder indicates the order in which runtime data is written to the corresponding physical storage location, and this writing order is determined based on the logical segment identifier; Step 4-4: After completing the operation of writing the running data to the database server, the monitoring center performs an integrity verification operation on the written running data to confirm that the running data has not been lost, truncated or misaligned during the writing process. Steps 4-5: After the integrity verification operation is successful, the monitoring center updates the status information of the corresponding logical storage address in the storage location mapping table, which is used to indicate that the running data under this logical location has been written and can be queried and called later.

[0013] Preferably, step 5 includes the following steps: Step 5-1: After the client sends a data query request to the monitoring center, the monitoring center extracts query conditions related to the inverter's operating data from the data query request. The query conditions include at least the device identifier, data type, and time range. Step 5-2: The monitoring center converts the extracted query conditions into corresponding logical storage addresses, as shown below: LogicalAddressSet=g(DeviceID, DataType, TimeRange); Wherein, LogicalAddressSet represents the logical storage address, DeviceID represents the inverter identifier, DataType represents the data type, and TimeRange represents the time range; g() represents the rule for generating a set of logical storage addresses corresponding to the query conditions based on the device identifier, data type, and time range in the query conditions; Step 5-3: The monitoring center uses the logical storage address LogicalAddressSet to find the corresponding storage location mapping and resolves the physical storage location associated with the logical storage address LogicalAddressSet and its write status information. PhysicalLocationSet=Map(LogicalAddressSet); Where PhysicalLocationSet represents the physical storage location; Map() represents the lookup rule that maps logical storage addresses to corresponding physical storage locations through a storage location mapping table; Step 5-4: After determining the physical storage location (PhysicalLocationSet), the monitoring center reads the corresponding runtime data in an ordered manner according to the logical storage address (LogicalAddressSet) and the logical segment identifier (SegmentID). The read runtime data is then logically reassembled to restore the continuity of the runtime data in both the time and device dimensions. The logical reassembly is represented as follows: ReadOrder ∝ SegmentID; Where ReadOrder represents the logical reorganization order of the data during the reading process; Step 5-5: The monitoring center encapsulates the logically reorganized operational data and sends it to the client for display.

[0014] This invention provides a real-time acquisition and storage method for key operating parameters of frequency converters. It solves the technical problem of achieving complete acquisition, stable storage, and rapid location and query of operating data from multiple frequency converters under conditions of high concurrency and out-of-order upload. By setting up acquisition units with local storage capabilities on the frequency converter side, the invention continuously stores operating data in time-ordered format and uploads it in data blocks when preset conditions are met. This ensures the integrity of the original operating data while effectively reducing network transmission pressure. By constructing a storage location mechanism independent of the physical storage location at the monitoring center, logical storage addresses are generated based on time intervals, device identifiers, and data types. This achieves unified management of out-of-order, high-concurrency data arrivals, avoiding storage chaos caused by uncertain data arrival order. By mapping logical storage addresses to physical storage locations, rapid location and orderly reorganization can be completed during data writing and querying solely based on logical location information. This improves the storage and query efficiency of large-scale frequency converter operating data and is suitable for data acquisition scenarios involving multiple devices and long-term continuous operation. Attached Figure Description

[0015] Figure 1 This is the main flowchart of the present invention; Figure 2 This is a flowchart of edge-side operation data acquisition and time alignment according to the present invention; Figure 3 This is a flowchart of edge-side data splitting and data packet encapsulation according to the present invention; Figure 4 This is a flowchart of the central-side logical segmentation and logical storage address generation of the present invention; Figure 5 This is a flowchart illustrating the mapping and writing of logical addresses to physical storage locations according to the present invention; Figure 6 This is a flowchart of the data query and reorganization based on the storage location mechanism of the present invention. Detailed Implementation

[0016] Depend on Figures 1-6 The method for real-time acquisition and storage of key operating parameters of a frequency converter, as shown, includes the following steps: Step 1: The acquisition unit set on the inverter side collects the operating data of the inverter in real time during operation. Each acquisition unit contains an edge-side storage unit.

[0017] Each acquisition unit is an edge node, and each edge node is responsible for acquiring the operating data of one frequency converter. The types of operating data include current, voltage, and frequency. The acquisition unit consists of an internal sampling module of the frequency converter or a sensor module connected to it, and a data processing module; the edge-side storage unit is the local memory of the acquisition unit.

[0018] The acquisition unit sorts the operating data by time according to the dimension of a single frequency converter, and then stores it in the edge storage unit; the acquisition unit divides the operating data in the edge storage unit into blocks according to time intervals, generates multiple data blocks, and adds source information to each data block; Step 1 includes the following steps: Step 1-1: During the operation of the frequency converter, the edge nodes continuously collect the operating data of the frequency converter and uniformly time-stamp the collected operating data to form a raw data sequence with time consistency. i (t k Specifically, it is expressed as: Data i (t k )={I(t k ),U(t k ),f(t k ),…}; Among them, t k Indicates the time; I, U, and f represent current, voltage, and frequency, respectively; i represents the inverter number. Step 1-2: Convert the original data sequence Data i (t k When writing to edge-side storage cells, the write order must satisfy the following conditions: Data i (t1)→Data i (t2)→⋯→Data i (t n ), t1<t2<...<t n ; Where n represents the last moment; Steps 1-3: According to a preset fixed time interval, the continuously written running data is segmented to generate multiple consecutive data blocks; Steps 1-4: Add corresponding metadata to each data block. i,m Specifically, it is expressed as follows: Meta i,m ={DeviceID,DataTypeSet,T start ,T end ,SeqNo}; Where DeviceID represents the inverter identifier; DataTypeSet represents the collection of all types of operating data contained in this data block; T start Indicates the start time of data collection; T end This indicates the end time of data collection; SeqNo indicates the sequence number of the locally generated sequence.

[0019] Step 2: When the preset triggering conditions are met, the acquisition unit selects at least one complete data block from the edge storage unit as the object to be uploaded, and splits the running data contained in the selected data block according to the data type. The running data of the same data type in the same data block are reassembled into the corresponding edge data packets. The acquisition unit uploads the edge data packets to the monitoring center asynchronously through the network. By splitting runtime data of different data types in the same data block at the edge and encapsulating them into edge-side data packets, the monitoring center can independently generate logical storage addresses based on data type dimension in subsequent steps. This avoids storage conflicts and increased query complexity caused by mixing runtime data with different sampling frequencies or data structures on the center side.

[0020] Step 2 includes the following steps: Step 2-1: When the preset triggering conditions are met, the edge node selects one or more complete data blocks from the edge-side storage unit as the objects to be uploaded; The preset trigger conditions are: triggering at a fixed time interval, triggering when the cumulative data volume reaches a threshold, or triggering when the local cache capacity reaches a threshold. Step 2-2: For each selected data block, split the running data contained therein according to the type of running data; extract the running data of the same data type in the same data block to form the corresponding data type subsequence; Steps 2-3: For each data type subsequence, combine it with the metadata of the original data block to encapsulate and generate the corresponding edge-side data packet. The edge-side data packet contains the running data sequence of a single data type, the inverter identifier DeviceID, the data type DataType, and the data acquisition start time T. start Data collection end time T end and the local sequence number SeqNo; Steps 2-4: The acquisition unit will asynchronously upload multiple edge-side data packets to the monitoring center via the network. Edge-side data packets from different frequency converters, different data types, or different time intervals will arrive at the monitoring center in an out-of-order or concurrent manner.

[0021] Step 3: The monitoring center receives and parses the received edge-side data packets, extracts the metadata associated with the running data, generates logical storage addresses based on the metadata, establishes a storage location mapping between the logical storage addresses and data metadata, and generates a data logical location structure independent of the physical storage location. Step 3 includes the following steps: Step 3-1: The monitoring center receives edge-side data packets uploaded from multiple edge nodes, and parses each edge-side data packet. Specifically, it separates the running data from the metadata describing the source and time of the running data, and marks the frequency converter, acquisition time range and data type corresponding to the edge-side data packet. Step 3-2: Based on the data collection start time contained in the edge-side data packets, the monitoring center maps the edge-side data packets to a preset time interval, so that the data on the continuous time axis is logically divided into multiple time periods. Specifically, a time interval length ΔT is preset, and the continuous time axis is discretized into time buckets of fixed length. The time buckets are represented as follows: TimeBucket=⌊T star / ΔT⌋; Among them, T star Indicates the start time of data collection, specifically obtained from metadata; TimeBucket is the time bucket. Step 3-3: Within the same time interval, for edge-side data packets originating from the same frequency converter and belonging to the same data type, the monitoring center assigns a logical segment identifier to the edge-side data packet to describe the logical sequence relationship of the edge-side data packet among similar data, specifically as follows: SegmentID=SeqID or AutoIncrement(TimeBucket, DeviceID, DataType) Wherein, DeviceID represents the inverter identifier, DataType represents the data type, SeqID represents the sequence number of the edge node, SegmentID represents the logical segment identifier, and AutoIncrement() represents the automatic increment function, which specifically assigns an incrementing logical segment identifier to different data segments in sequence under the same time interval, device identifier, and data type conditions. In this embodiment, AutoIncrement() specifically executes, within the same logical time bucket, the same device, and the same data type range, according to the edge-side generated sequence number SeqID or the arrival order, to assign a monotonically increasing logical sequence number to the data segment.

[0022] The generation rule for the logical segment identifier does not depend on the physical auto-increment field of the database, but is a logical sequence numbering rule that is independently maintained based on the logical time interval, device identifier, and data type dimension.

[0023] DeviceID, DataType, and SeqID are all obtained from metadata; Steps 3-4: The monitoring center concatenates the time bucket, inverter identifier, data type, and logical segment identifier to generate a unique logical storage address. This logical storage address is used to describe only the logical location of the running data in the edge-side data packet under the [time + device + type] dimension. The logical storage address is represented as follows: LogicalAddress=<TimeBucket,DeviceID,DataType,SegmentID> ; Wherein, LogicalAddress is the logical storage address; Steps 3-5: The monitoring center associates the generated logical storage addresses with the corresponding runtime data metadata and writes them into a storage location mapping, which serves as the basic index information for data writing and querying. The storage location mapping is represented as follows: MappingRecord={LogicalAddress, DeviceID, DataType, T start T end ,Fs,N} Where MappingRecord is the storage location mapping, T end T is the end time of data acquisition, Fs is the sampling frequency, and N is the number of data points; end Both Fs and N are obtained from the metadata; Steps 3-6: The monitoring center establishes a storage location mapping list based on the storage location mapping, and constructs a data logical location structure independent of the physical storage location; The physical storage location refers to the physical storage address location of the database server.

[0024] Step 4: The monitoring center writes the corresponding running data into the database server according to the logical storage address and storage location mapping. When performing the write operation, it supports out-of-order writing and concurrent writing, and updates the storage status information after the writing is completed. Step 4 includes the following steps: Step 4-1: The monitoring center uses the logical storage address and storage location mapping as the sole basis for data writing, triggering the data writing process. Specifically, based on the time interval identifier, device identifier, and data type in the logical storage address, the running data to be written is mapped to the corresponding storage partition in the database server, thus associating the logical storage structure with the physical storage structure to determine the physical storage location. This is illustrated as follows: PhysicalLocation=f(TimeBucket, DeviceID, DataType); Wherein, PhysicalLocation represents the association between logical storage structure and physical storage structure; TimeBucket is the time bucket; DeviceID represents the inverter identifier; DataType represents the data type; f() represents the mapping function, specifically a mapping rule for selecting the physical storage location of running data based on the time interval identifier, device identifier, and data type in the logical storage address; in this embodiment, the f() mapping relationship is as follows: Time → Time Partition.

[0025] Equipment → Equipment List / Equipment Catalog.

[0026] Type → Column Family / Table.

[0027] The partitioning in this embodiment is based on logical storage addresses and is used to support out-of-order writing and subsequent fast location.

[0028] The process of determining the physical storage location relies only on the abstract dimension information in the logical storage address, and does not depend on the arrival order or writing order of edge-side data packets, thereby supporting stable storage of running data under out-of-order and concurrent writing conditions.

[0029] Step 4-2: After determining the physical storage location, organize the running data under the same logical storage address sequentially according to the logical segment identifier (SegmentID), and write the running data into the corresponding storage space in sequence. The sequential organization is represented as follows: WriteOrder ∝ SegmentID; WriteOrder indicates the order in which runtime data is written to the corresponding physical storage location, and this writing order is determined based on the logical segment identifier; In this embodiment, SegmentID has already guaranteed the conditions of the same time bucket, the same device identifier, and the same data type. Writing in order of SegmentID can restore logical continuity under the condition of out-of-order arrival.

[0030] Step 4-4: After completing the operation of writing the running data to the database server, the monitoring center performs an integrity verification operation on the written running data to confirm that the running data has not been lost, truncated or misaligned during the writing process. Steps 4-5: After the integrity verification operation is successful, the monitoring center updates the status information of the corresponding logical storage address in the storage location mapping table, which is used to indicate that the running data under this logical location has been written and can be queried and called later.

[0031] Step 5: After the client sends a data query request to the monitoring center, the monitoring center filters and locates the logical storage address according to the storage location mapping, and reads the corresponding operating data from the database server according to the filtering and location results, thus completing the fast query of the operating data of the specified frequency converter.

[0032] Step 5 includes the following steps: Step 5-1: After the client sends a data query request to the monitoring center, the monitoring center extracts query conditions related to the inverter's operating data from the data query request. The query conditions include at least the device identifier, data type, and time range. Step 5-2: The monitoring center converts the extracted query conditions into corresponding logical storage addresses, as shown below: LogicalAddressSet=g(DeviceID, DataType, TimeRange); Wherein, LogicalAddressSet represents the logical storage address, DeviceID represents the inverter identifier, DataType represents the data type, and TimeRange represents the time range; g() represents the rule for generating a set of logical storage addresses corresponding to the query conditions based on the device identifier, data type, and time range in the query conditions; In this embodiment, g() specifically calculates one or more time buckets covering the time range based on the device identifier, data type, and time range in the query conditions, and combines them to generate a corresponding set of logical storage addresses.

[0033] Step 5-3: The monitoring center uses the logical storage address LogicalAddressSet to find the corresponding storage location mapping and resolves the physical storage location associated with the logical storage address LogicalAddressSet and its write status information. PhysicalLocationSet=Map(LogicalAddressSet); Where PhysicalLocationSet represents the physical storage location; Map() represents the lookup rule that maps logical storage addresses to corresponding physical storage locations through a storage location mapping table; In this embodiment, the logical storage address is only used to describe the logical location of running data in the time dimension, device dimension, and data type dimension, and does not directly correspond to the physical storage address in the database server.

[0034] The monitoring center maintains a storage location mapping table in advance. This table uses logical storage addresses as index keys and records the physical storage locations corresponding to those logical storage addresses and their write status information. When it is necessary to read or write operational data, the monitoring center does not generate physical storage addresses through calculation. Instead, it performs a lookup operation in the storage location mapping table based on the logical storage address to resolve the set of physical storage locations associated with the logical storage address.

[0035] Map() represents a logical address resolution process based on a storage location mapping table, used to convert a set of logical storage addresses into a set of corresponding physical storage locations.

[0036] Map() = parsing rules for logical address → storage location record → physical storage location set.

[0037] Specifically, it is expressed as follows: LogicalAddressSet is a set of logical address keys, such as<TimeBucket, DeviceID, DataType,SegmentID> ; The monitoring center has already maintained a "storage location mapping table", and each record is as follows: MappingRecord = { LogicalAddress Physical Location WriteStatus, OffsetInfo } Map() is used to "store the location mapping table" using logical storage addresses.

[0038] Step 5-4: After determining the physical storage location (PhysicalLocationSet), the monitoring center reads the corresponding runtime data in an ordered manner according to the logical storage address (LogicalAddressSet) and the logical segment identifier (SegmentID). The read runtime data is then logically reassembled to restore the continuity of the runtime data in both the time and device dimensions. The logical reassembly is represented as follows: ReadOrder ∝ SegmentID; Wherein, ReadOrder represents the logical reorganization order of the running data during the reading process; the logical reorganization order is determined by sorting according to the size relationship of the logical segment identifiers.

[0039] Step 5-5: The monitoring center encapsulates the logically reorganized operational data and sends it to the client for display.

[0040] This invention provides a real-time acquisition and storage method for key operating parameters of frequency converters. It solves the technical problem of achieving complete acquisition, stable storage, and rapid location and query of operating data from multiple frequency converters under conditions of high concurrency and out-of-order upload. By setting up acquisition units with local storage capabilities on the frequency converter side, the invention continuously stores operating data in time-ordered format and uploads it in data blocks when preset conditions are met. This ensures the integrity of the original operating data while effectively reducing network transmission pressure. By constructing a storage location mechanism independent of the physical storage location at the monitoring center, logical storage addresses are generated based on time intervals, device identifiers, and data types. This achieves unified management of out-of-order, high-concurrency data arrivals, avoiding storage chaos caused by uncertain data arrival order. By mapping logical storage addresses to physical storage locations, rapid location and orderly reorganization can be completed during data writing and querying solely based on logical location information. This improves the storage and query efficiency of large-scale frequency converter operating data and is suitable for data acquisition scenarios involving multiple devices and long-term continuous operation.

Claims

1. A method for real-time acquisition and storage of key operating parameters of a frequency converter, characterized in that: Includes the following steps: Step 1: The acquisition unit set up on the inverter side collects the operating data in real time during the operation of the inverter. Each acquisition unit contains an edge storage unit. The acquisition unit sorts the operating data by time according to the dimension of a single inverter and then stores it in the edge storage unit. The acquisition unit divides the operating data in the edge storage unit into blocks according to time intervals, generates multiple data blocks, and adds source information to each data block. Step 2: When the preset triggering conditions are met, the acquisition unit selects at least one complete data block from the edge storage unit as the object to be uploaded, and splits the running data contained in the selected data block according to the data type. The running data of the same data type in the same data block are reassembled into the corresponding edge data packets. The acquisition unit uploads the edge data packets to the monitoring center asynchronously through the network. Step 3: The monitoring center receives and parses the received edge-side data packets, extracts the metadata associated with the running data, generates logical storage addresses based on the metadata, establishes a storage location mapping between the logical storage addresses and data metadata, and generates a data logical location structure independent of the physical storage location. Step 4: The monitoring center writes the corresponding running data into the database server according to the logical storage address and storage location mapping, and updates the storage status information after the writing is completed; Step 5: After the client sends a data query request to the monitoring center, the monitoring center filters and locates the logical storage address according to the storage location mapping, and reads the corresponding operating data from the database server according to the filtering and location results, thus completing the fast query of the operating data of the specified frequency converter.

2. The method for real-time acquisition and storage of key operating parameters of a frequency converter as described in claim 1, characterized in that: Each acquisition unit is an edge node, and each edge node is responsible for acquiring the operating data of one frequency converter. The types of operating data include current, voltage, and frequency. The acquisition unit consists of an internal sampling module of the frequency converter or a sensor module connected to it, and a data processing module; the edge-side storage unit is the local memory of the acquisition unit.

3. The method for real-time acquisition and storage of key operating parameters of a frequency converter as described in claim 2, characterized in that: Step 1 includes the following steps: Step 1-1: During the operation of the frequency converter, the edge nodes continuously collect the operating data of the frequency converter and uniformly time-stamp the collected operating data to form a raw data sequence with time consistency. i (t k Specifically, it is expressed as: Data i (t k )={I(t k ),U(t k ),f(t k ),…}; Among them, t k Indicates the time; I, U, and f represent current, voltage, and frequency, respectively; i represents the inverter number. Step 1-2: Convert the original data sequence Data i (t k When writing to edge-side storage cells, the write order must satisfy the following conditions: Data i (t1)→Data i (t2)→⋯→Data i (t n ),t1<t2<...<t n ; Where n represents the last moment; Steps 1-3: According to a preset fixed time interval, the continuously written running data is segmented to generate multiple consecutive data blocks; Steps 1-4: Add corresponding metadata to each data block. i,m Specifically, it is expressed as follows: Meta i,m ={DeviceID,DataTypeSet,T start ,T end ,SeqNo}; Where DeviceID represents the inverter identifier; DataTypeSet represents the collection of all types of operating data contained in this data block; T start Indicates the start time of data collection; T end This indicates the end time of data collection; SeqNo indicates the sequence number of the locally generated sequence.

4. The method for real-time acquisition and storage of key operating parameters of a frequency converter as described in claim 3, characterized in that: Step 2 includes the following steps: Step 2-1: When the preset triggering conditions are met, the edge node selects one or more complete data blocks from the edge-side storage unit as the objects to be uploaded; The preset trigger conditions are: triggering at a fixed time interval, triggering when the cumulative data volume reaches a threshold, or triggering when the local cache capacity reaches a threshold. Step 2-2: For each selected data block, split the running data contained therein according to the type of running data; extract the running data of the same data type in the same data block to form the corresponding data type subsequence; Steps 2-3: For each data type subsequence, combine it with the metadata of the original data block to encapsulate and generate the corresponding edge-side data packet. The edge-side data packet contains the running data sequence of a single data type, the inverter identifier DeviceID, the data type DataType, and the data acquisition start time T. start Data collection end time T end and the local sequence number SeqNo; Steps 2-4: The acquisition unit will asynchronously upload multiple edge-side data packets to the monitoring center via the network. Edge-side data packets from different frequency converters, different data types, or different time intervals will arrive at the monitoring center in an out-of-order or concurrent manner.

5. The method for real-time acquisition and storage of key operating parameters of a frequency converter as described in claim 2, characterized in that: Step 3 includes the following steps: Step 3-1: The monitoring center receives edge-side data packets uploaded from multiple edge nodes, and parses each edge-side data packet. Specifically, it separates the running data from the metadata describing the source and time of the running data, and marks the frequency converter, acquisition time range and data type corresponding to the edge-side data packet. Step 3-2: Based on the data collection start time contained in the edge-side data packets, the monitoring center maps the edge-side data packets to a preset time interval, so that the data on the continuous time axis is logically divided into multiple time periods. Specifically, a time interval length ΔT is preset, and the continuous time axis is discretized into time buckets of fixed length. The time buckets are represented as follows: TimeBucket=⌊T star / ΔT⌋; Among them, T star Indicates the start time of data collection, specifically obtained from metadata; TimeBucket is the time bucket. Step 3-3: Within the same time interval, for edge-side data packets originating from the same frequency converter and belonging to the same data type, the monitoring center assigns a logical segment identifier to the edge-side data packet to describe the logical sequence relationship of the edge-side data packet among similar data, specifically as follows: SegmentID=SeqID or AutoIncrement(TimeBucket, DeviceID, DataType) Where DeviceID represents the inverter identifier, DataType represents the data type, SeqID represents the sequence number of the edge node, SegmentID represents the logical segment identifier, and AutoIncrement() represents the automatic increment function, which is a sequential numbering rule for assigning incrementing logical segment identifiers to different data segments under the same time interval, device identifier, and data type conditions. DeviceID, DataType, and SeqID are all obtained from metadata; Steps 3-4: The monitoring center concatenates the time bucket, inverter identifier, data type, and logical segment identifier to generate a unique logical storage address. This logical storage address is used to describe only the logical location of the running data in the edge-side data packet under the [time + device + type] dimension. The logical storage address is represented as follows: LogicalAddress=<TimeBucket,DeviceID,DataType,SegmentID> ; Wherein, LogicalAddress is the logical storage address; Steps 3-5: The monitoring center associates the generated logical storage addresses with the corresponding runtime data metadata and writes them into a storage location mapping, which serves as the basic index information for data writing and querying. The storage location mapping is represented as follows: MappingRecord={LogicalAddress,DeviceID,DataType,T start ,T end ,Fs,N} Where MappingRecord is the storage location mapping, T end T is the end time of data acquisition, Fs is the sampling frequency, and N is the number of data points; end Both Fs and N are obtained from the metadata; Steps 3-6: The monitoring center establishes a storage location mapping list based on the storage location mapping, and constructs a data logical location structure independent of the physical storage location; The physical storage location refers to the physical storage address location of the database server.

6. The method for real-time acquisition and storage of key operating parameters of a frequency converter as described in claim 5, characterized in that: Step 4 includes the following steps: Step 4-1: The monitoring center uses the logical storage address and storage location mapping as the sole basis for data writing, triggering the data writing process. Specifically, based on the time interval identifier, device identifier, and data type in the logical storage address, the running data to be written is mapped to the corresponding storage partition in the database server, thus associating the logical storage structure with the physical storage structure to determine the physical storage location. This is illustrated as follows: PhysicalLocation=f(TimeBucket, DeviceID, DataType); Wherein, PhysicalLocation represents the association between logical storage structure and physical storage structure; TimeBucket is the time bucket; DeviceID represents the inverter identifier; DataType represents the data type; f() represents the mapping function, which is a mapping rule for selecting the physical storage location of running data based on the time interval identifier, device identifier and data type in the logical storage address; Step 4-2: After determining the physical storage location, organize the running data under the same logical storage address sequentially according to the logical segment identifier (SegmentID), and write the running data into the corresponding storage space in sequence. The sequential organization is represented as follows: WriteOrder ∝ SegmentID; WriteOrder indicates the order in which runtime data is written to the corresponding physical storage location, and this writing order is determined based on the logical segment identifier; Step 4-4: After completing the operation of writing the running data to the database server, the monitoring center performs an integrity verification operation on the written running data to confirm that the running data has not been lost, truncated or misaligned during the writing process. Steps 4-5: After the integrity verification operation is successful, the monitoring center updates the status information of the corresponding logical storage address in the storage location mapping table, which is used to indicate that the running data under this logical location has been written and can be queried and called later.

7. The method for real-time acquisition and storage of key operating parameters of a frequency converter as described in claim 6, characterized in that: Step 5 includes the following steps: Step 5-1: After the client sends a data query request to the monitoring center, the monitoring center extracts query conditions related to the inverter's operating data from the data query request. The query conditions include at least the device identifier, data type, and time range. Step 5-2: The monitoring center converts the extracted query conditions into corresponding logical storage addresses, as shown below: LogicalAddressSet=g(DeviceID, DataType, TimeRange); Wherein, LogicalAddressSet represents the logical storage address, DeviceID represents the inverter identifier, DataType represents the data type, and TimeRange represents the time range; g() represents the rule for generating a set of logical storage addresses corresponding to the query conditions based on the device identifier, data type, and time range in the query conditions; Step 5-3: The monitoring center uses the logical storage address LogicalAddressSet to find the corresponding storage location mapping and resolves the physical storage location associated with the logical storage address LogicalAddressSet and its write status information. PhysicalLocationSet=Map(LogicalAddressSet); Where PhysicalLocationSet represents the physical storage location; Map() represents the lookup rule that maps logical storage addresses to corresponding physical storage locations through a storage location mapping table; Step 5-4: After determining the physical storage location (PhysicalLocationSet), the monitoring center reads the corresponding runtime data in an ordered manner according to the logical storage address (LogicalAddressSet) and the logical segment identifier (SegmentID). The read runtime data is then logically reassembled to restore the continuity of the runtime data in both the time and device dimensions. The logical reassembly is represented as follows: ReadOrder ∝ SegmentID; Where ReadOrder represents the logical reorganization order of the data during the reading process; Step 5-5: The monitoring center encapsulates the logically reorganized operational data and sends it to the client for display.

Images