An entity state bias quantization encoding and decoding method and system based on a standardized benchmark set

Abstract

The application discloses an entity state deviation quantization encoding and decoding method and system based on a standardized benchmark set, and belongs to the technical field of data encoding and compression. The method establishes a standardized benchmark set containing multi-dimensional standard state values for each entity type; current state data of the entity is collected, and the deviation from the benchmark value is calculated; the deviation value is quantization encoded according to the independently preset quantization precision of the dimension to generate a deviation vector; the deviation vector is stored in association with an entity identifier and a time stamp, including storing a zero-value deviation vector when the state is unchanged, instead of storing complete state; when queried, the deviation vector and the benchmark set are read, the quantization deviation value is superimposed on the benchmark value to decode and restore approximate complete state data within the quantization precision range, and the decoding does not depend on any previous time data and can be independently decoded at any time. According to the application, only the state deviation is stored to replace the complete state storage, the storage space is reduced by more than 65%, the query time complexity is O(1), the encoding precision can be independently controlled according to the dimension, and the application can be widely applied to the scenes of Internet of Things device monitoring, digital twin data synchronization, wearable device health monitoring and the like.

Description

Technical Field

[0001] This invention relates to the field of data encoding and data compression technology, specifically to a method and system for encoding and decoding entity state deviation quantization based on a standardized reference set, and in particular to an encoding method that supports independent random access under lossy compression. Background Technology

[0002] With the rapid development of IoT, big data, and digital twin technologies, systems need to collect, store, and transmit massive amounts of entity state data in real time. In existing technologies, the storage and transmission of entity state data typically employs a complete state snapshot approach, meaning that after each acquisition of the latest entity state, all state data is encoded, stored, or transmitted. This method has the following technical drawbacks: First, it suffers from high data redundancy and significant storage overhead. Entity state data typically exhibits strong temporal correlation in a time series, with minimal differences between adjacent time points. The full state snapshot approach fails to utilize this data characteristic, resulting in the repeated storage of large amounts of duplicate data. Second, network transmission bandwidth consumption is high. In IoT and edge computing scenarios, the frequent reporting of entity status data consumes a large amount of network bandwidth. Taking smart meters as an example, a complete status data report (including dozens of fields such as voltage, current, power factor, and cumulative power consumption) is submitted every 15 minutes. A single record can reach several KB, and tens of thousands of devices report tens of GB of data daily. Third, status query response time is long. In storage systems containing massive amounts of entity status data, querying the complete status at a specific moment requires reading all field data, resulting in high disk I / O overhead and long response time. Fourth, there is a lack of differentiated control over coding precision. Different dimensions of state data have different precision requirements. For example, in smart grid monitoring, voltage data requires high precision (error <0.5%), while power factor data has relatively lower precision requirements. Existing solutions use a uniform coding precision, failing to achieve a fine balance between storage efficiency and data accuracy. Fifth, the coupling between state data and baseline data is high. In existing solutions, entity type characteristic data and dynamically changing data are stored together, which means that static baseline information and dynamically changing information must be processed simultaneously when updating data, increasing the complexity of data processing; There are some differential storage schemes in the existing technology, such as data compression methods based on forward differential or backward differential. However, these methods have two significant limitations: (1) the differential benchmark is not fixed and needs to depend on the preceding data step by step. When decoding, all historical differential data must be played back in sequence, and it is not possible to directly jump to any time for querying; (2) there is a lack of a standardized benchmark set definition and management mechanism, and it is impossible to establish a unified and reusable reference benchmark for the same entity type; Therefore, there is an urgent need for an encoding and decoding method that can utilize the temporal correlation of entity states, support dimensionally differentiated precision control, and support independent querying of states at any time, in order to solve the above-mentioned technical problems. Summary of the Invention

[0003] (a) Purpose of the invention This invention provides a method and system for quantitative encoding and decoding of entity state deviation based on a standardized reference set, aiming to solve the technical problems in the prior art, such as high data redundancy in entity state data storage, large transmission bandwidth consumption, slow query response, coarse precision control, and non-fixed differential references that depend on preceding data.

[0004] (II) Technical Solution This invention is achieved using the following technical solution: A method for quantizing and decoding entity state deviations based on a standardized reference set, characterized by the following steps: Step S1: Construct a standardized benchmark set; A standardized benchmark set is established for each entity type, which contains the standard state values ​​of that entity type in multiple dimensions; The entity type refers to an entity category with the same attribute structure, such as "smart meter device", "user health record", "vehicle status record", etc. The multiple dimensions refer to the various attribute dimensions that characterize the state of an entity, including but not limited to physical dimension (voltage, current, temperature, speed, etc.), state attribute dimension (online status, operating mode, fault code, etc.), and cumulative dimension (cumulative power consumption, cumulative runtime, etc.). The standard state value is the typical state value of this entity type under standard operating conditions, and is determined by any of the following methods: Method A: Perform statistical clustering on the historical state data of all entities of the same type, and use the cluster centers as the standard state values; Method B: Set according to industry standards or the equipment's factory specifications; Method C: Configured by the administrator according to the actual business scenario; Each standardized benchmark set is assigned a unique benchmark set identifier, which is then associated with and stored in relation to the corresponding entity type code. Step S2: Collect current status data and calculate the deviation value; Collect the current state data of the entity, compare the current state data with the standard state values ​​of the corresponding dimensions in the standardized benchmark set, and calculate the deviation value for each dimension; The deviation value is calculated as follows: Deviation value = Current state value - Standard state value or Deviation value = (Current state value - Standard state value) / Standard state value; the two calculation methods can be selected according to the dimension type. For dimensions with explicit dimensions, the difference method is used; for relative value dimensions or ratio dimensions, the relative deviation method is used. Step S3: Deviation quantization encoding generates deviation vector: The deviation values ​​of each dimension are quantized and encoded according to a preset quantization precision to generate a deviation vector; The quantization precision refers to the number of bits used to encode the deviation value or the quantization step size. Higher quantization precision results in higher precision of the encoded data, but also requires more storage space. The quantization precision for each dimension can be set independently and dynamically adjusted according to the dimension's importance. For critical dimensions (such as temperature in security monitoring or heart rate in medical monitoring), high quantization precision (e.g., 16-bit encoding) is used; for secondary dimensions (such as ambient brightness or auxiliary status indicators), low quantization precision (e.g., 4-bit encoding or 1-bit flag) is used. The deviation vector uses a bit-field compressed storage method, which compactly arranges the deviation codes of multiple dimensions in a continuous binary bit field according to a predefined bit width, eliminating storage fragmentation caused by padding bytes between fields; Step S4: Associate and store the deviation vector: The deviation vector is associated with and stored with the entity's unique identifier and timestamp, instead of storing complete entity state data; When the deviation of an entity's current state from the standardized reference set is zero in all dimensions, or the change in deviation in each dimension is less than a preset minimum change threshold, the deviation vector is generated as a zero-value deviation vector with zero deviation encoding in each dimension, and is still stored to ensure that there is an independently decodeable deviation record at any given time. The zero-value deviation vector can use a preset zero-value compact encoding format to further compress storage space. Step S5: Decode and restore near-complete state data: In response to an entity status query request, the deviation vector and the standardized benchmark set are read, and the quantized deviation values ​​of each dimension in the deviation vector are superimposed on the standard status values ​​of the corresponding dimensions in the standardized benchmark set. The entity's approximately complete status data at the time of query, within the allowable error range of quantization accuracy, is then decoded and restored. The calculation method for the decoding and restoration is as follows: Approximate value of entity state = Standard state value + Quantization deviation restoration value The decoding process does not rely on any previous state data; the deviation vector at any time can be decoded independently, and it supports direct jump to any time for querying. The method also includes the following preferred technical solutions: Step S6: Dynamic update of the benchmark set: When the standardized benchmark set for an entity type is updated (e.g., due to equipment upgrades or changes in operating conditions requiring adjustments to standard state values), the system automatically updates the benchmark set version number and associates the newly acquired deviation vector with the new version of the benchmark set; Step S7: Deviation Vector Time Series Management: Maintain a time-series index of the deviation vector for each entity, supporting querying historical versions of the deviation vector by time range, and exporting the entity's state change trajectory by time range; Step S8: Conservation Verification: For dimensions with total quantity conservation constraints, and when the entity is divided into a hierarchical structure containing a parent and several child entities, conservation verification is performed after decoding and restoration: the state values ​​of each child entity are summarized and compared with the total quantity of the parent entity. If the deviation exceeds a preset threshold, a consistency alarm is triggered.

[0005] (III) Beneficial Effects Compared with the prior art, the present invention has the following advantages: 1. Significantly reduced storage space. Instead of storing the complete state, only the deviation values ​​relative to the standardized reference set are stored, and these deviation values ​​are quantized and stored using bit-field compression. Since the deviation values ​​of entity states relative to the reference set are small in most cases, the number of bits required for deviation encoding is much smaller than that required for the complete state data. For example, with 32-dimensional state data, complete storage requires 32 bits per dimension (1024 bits total), while deviation encoding only requires 4-8 bits per dimension (128-256 bits total), reducing storage space by approximately 75%-87.5%. Even if the state remains unchanged, storing the zero-value compactly encoded deviation vector is still much smaller than storing the complete state data. 2. Significantly reduces network transmission bandwidth consumption. When IoT devices report status data, only offset-coded data packets need to be transmitted instead of complete status data packets, reducing the amount of data transmitted by more than 75%, effectively reducing the power consumption and communication costs of wireless communication modules; 3. Independent decoding, supporting direct query at any time. This invention uses a fixed standardized benchmark set as the decoding benchmark, and stores the corresponding deviation vector (including the zero value vector when the state remains unchanged) at each sampling time. This ensures that when querying the entity state at any time, the deviation vector at that time can be directly superimposed on the benchmark set for decoding without relying on any prior data. This avoids the defect of traditional differential coding schemes that require sequential replay of all historical data, and the query time complexity is O(1). 4. Supports independent control of encoding precision by dimension. Different quantization precisions can be used for deviation values ​​in different dimensions. High-precision encoding of key dimensions ensures data accuracy, while low-precision encoding of secondary dimensions saves storage space, achieving a fine balance between storage efficiency and data accuracy; 5. Decoupling of baseline sets and deviation vectors. Standardized baseline sets for entity types are stored and version-managed independently. Status updates only involve changes to deviation vectors; updating the baseline set does not affect the decodeability of previously stored historical deviation vectors (version number management is required). 6. Technological Synergy with Existing Patent Systems. This invention can supplement the data encoding layer of the patent application "A Closed-Loop Method for Quantitative Evaluation and Financial Recording of Subjects and Objects Based on Multidimensional Value Accounting" (A8), providing an efficient encoding and transmission scheme for real-time synchronization of state data in five-dimensional value accounting. Simultaneously, the conservation verification step in this invention can work in conjunction with the parent increment conservation scheme (the specific implementation of A1) to form a complete "deviation quantification encoding → efficient transmission → conservation verification" data link. Attached Figure Description Figure 1 This is a flowchart illustrating the overall process of the method of the present invention. Figure 2 This is a schematic diagram of the standardized benchmark set data structure of this invention; Figure 3 This is a schematic diagram of the bit-field compressed storage structure of the deviation vector of the present invention; Figure 4 This is a timing diagram of the deviation vector generation and storage in this invention; Figure 5 This is a flowchart illustrating the decoding and restoration process for approximately complete state data in this invention. Figure 6 This is a system module architecture diagram of the present invention; The same reference numerals in each figure indicate the same or corresponding technical features. Detailed Implementation The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. These embodiments are for illustrative purposes only and should not be construed as limiting the scope of the invention. Example 1: Overall Method Flow Figure 1 Here is an overall flowchart of the method of the present invention: Figure 1 As shown, the method of the present invention includes the following core steps: Step S101, Construct a standardized benchmark set: Establish a standardized benchmark set for each entity type. The benchmark set contains the standard state values ​​B1, B2, ..., B_K of the entity type in K dimensions, and assign a benchmark set identifier BS_ID; Step S102, Collect current state data: Collect the current state values ​​C1, C2, ..., C_K of the entity through sensors, API interfaces, or manual input; Step S103, calculate the deviation value: calculate the deviation value Δᵢ = Cᵢ — Bᵢ (or relative deviation) for each dimension; Step S104, Quantization encoding: Based on the preset quantization precisions Q1, Q2, ..., Q_K for each dimension, the deviation value Δᵢ is encoded into a binary code stream Eᵢ; Step S105, Generate the deviation vector: Assemble the codes Eᵢ of each dimension into a deviation vector DV according to a predefined format. If all deviation values ​​are zero or the changes are within a threshold, a zero-value deviation vector is generated; Step S106, Associated Storage: Associate the deviation vector DV with the entity identifier E_ID and the timestamp T. The zero-value deviation vector is stored using compact encoding to ensure that there is an independently decodeable record at each time point. Step S107, Decoding and Restoration: In response to the query request, read the bias vector DV and the corresponding normalized reference set BS, calculate the quantization restoration value Rᵢ = Bᵢ + decode(Eᵢ), and obtain the approximately complete state within the quantization accuracy range.

[0006] Example 2: Data Structure of Standardized Benchmark Set Figure 2 This is a schematic diagram of the standardized benchmark set data structure of this invention. (See diagram below.) Figure 2 As shown, the data structure of the standardized benchmark set includes the following fields: Field Name Data Type Description BS_ID String(32) Unique identifier for the base set EntityType String(16) corresponds to the entity type code. Version Integer Base set version number DimCount (Integer) represents the number of dimensions (K). DimNames String[] List of names for each dimension BaseValues ​​Float[] is a list of standard state values ​​for each dimension. Precisions Integer[] List of quantization precision (bits) for each dimension UpdateTime (DateTime) is the last update time of the baseline set. Taking a smart meter entity type as an example, the data content of its standardized benchmark set is as follows: Field value BS_ID BS_0001 EntityType SmartMeter_TypeA Version 3 DimCount 6 DimNames [Voltage, Current, Power, Power Factor, Total Charge, Temperature] BaseValues ​​[220.0, 5.0, 1100.0, 0.95, 10000.0, 25.0] Precisions [12, 10, 12, 8, 16, 8].

[0007] Example 3: Bit-field compressed storage structure for deviation vector Figure 3 This is a schematic diagram of the bit-field compressed storage structure of the deviation vector in this invention. Figure 3 As shown, the bit-field compressed storage structure of the deviation vector adopts a compact, continuous bit-field arrangement: The deviation value for each dimension is quantized and encoded according to the preset quantization precision (number of bits) for that dimension. The encoded data for each dimension are then tightly arranged in a continuous binary bit field according to the dimension index, without any padding bytes between fields. Taking 6 dimensions as an example: Dimension 0 (Voltage): 12-bit precision, occupying bits [0:11] Dimension 1 (Current): 10-bit precision, occupying bits [12:21] Dimension 2 (Power): 12-bit precision, occupying bits [22:33] Dimension 3 (Power Factor): 8-bit precision, occupying bits [34:41] Dimension 4 (Cumulative Battery Charge): 16-bit precision, occupying bits [42:57] Dimension 5 (Temperature): 8-bit precision, occupying bits [58:65] The total space occupied is 66 bits, or 9 bytes (with less than an integer number aligned to the nearest whole byte). Storing a complete 6-dimensional floating-point number (32 bits per dimension) requires 192 bits (24 bytes). The compression ratio is 66:192≈1:2.91, reducing storage space by approximately 65.6%. Furthermore, if the deviation range of each dimension in the actual monitoring scenario is small (e.g., voltage deviation within ±5%), then a lower quantization precision (e.g., 8 bits for voltage, 6 bits for current) can be used, and the compression ratio can reach more than 1:4. For zero-value deviation vectors, a more compact predefined encoding (e.g., 1 byte flag bit) can be used, and the storage space can be further reduced.

[0008] Example 4: Timing of Deviation Vector Generation and Storage Figure 4 This is a timing diagram for the generation and storage of the deviation vector in this invention. (For example...) Figure 4 As shown, the system processes status data according to the following timing sequence: The sensor collects entity status data periodically (e.g., every 15 minutes). The system reads the standardized benchmark set corresponding to this entity type; The system calculates the deviation between the current state value and the reference value; The system generates a deviation vector using quantization encoding with preset precision; if the deviation in each dimension is zero or less than the threshold, a zero-value deviation vector is generated. The system associates and stores the deviation vector with the entity identifier and timestamp (zero-value deviation vectors are also stored); The system can optionally trigger subsequent processing: if the deviation in any dimension of the deviation vector exceeds the alarm threshold, an alarm event is generated; To verify the effectiveness of the technology, the applicant conducted comparative tests: Test metrics: Improvement of traditional complete storage solutions and the present invention's solution The amount of data reported in a single session has decreased from 1024 bits (32 dimensions × 32 bits) to 256 bits (32 dimensions × 8 bits average), a reduction of 75%. Average daily storage space (10,000 devices): Approximately 18.4GB; Approximately 4.6GB; a decrease of 75%. Status query response time is approximately 85ms, compared to approximately 12ms, representing an 85.3% improvement.

[0009] Example 5: Decoding and Restoring Nearly Complete State Data Flowchart Figure 5 This is a flowchart illustrating the decoding and restoration process of approximately complete state data in this invention. Figure 5 As shown, the decoding and restoration process includes the following steps: Step S501, Receive query request: The query request includes entity identifier E_ID and target time T; Step S502, retrieve the deviation vector: retrieve the deviation vector DV at the corresponding time based on E_ID and T (if the data stored at that time is zero-value compact encoding, then parse it into a zero-value deviation vector); Step S503, Obtain the benchmark set: Obtain the corresponding standardized benchmark set BS according to the entity type; Step S504, Dimensional Decoding: Decode the encoding of each dimension in the deviation vector to obtain the quantized deviation restoration value Δᵢ'; Step S505, Overlay the reference value: Overlay the quantization deviation restoration value onto the reference value: Rᵢ = Bᵢ + Δᵢ'; Step S506, Assemble the approximate complete state: Assemble the quantized values ​​R1, R2, ..., R_K of all dimensions into approximate complete state data within the quantization accuracy error range; Step S507: Return the query results.

[0010] Example 6: System Module Architecture Figure 6 This is a system module architecture diagram of the present invention. (For example...) Figure 6 As shown, the system of the present invention includes the following core modules: The benchmark set storage module (101) is configured to store standardized benchmark sets for each entity type, the benchmark sets containing standard state values ​​of the entity type in multiple dimensions; The deviation calculation module (102) is configured to receive the current state data of the entity, read the corresponding benchmark value in the benchmark set storage module, and calculate the deviation value between the current state data and the benchmark value; The encoding module (103) is configured to quantize and encode the deviation value according to the preset quantization precision of each dimension, and generate a deviation vector in bit-field compression format. The encoded data of each dimension in the deviation vector are arranged continuously in the binary bit field according to the dimension index order. When the deviation is zero or less than the threshold, a zero-value deviation vector is generated. The storage module (104) is configured to associate the deviation vector with the entity identifier and timestamp in the database, including the storage of the zero-value deviation vector, to ensure that there is an independently accessible deviation record at each sampling time; The decoding module (105) is configured to, in response to an entity state query request, read the deviation vector from the storage module, read the corresponding standardized reference set from the reference set storage module, superimpose the quantization deviation values ​​of each dimension in the deviation vector onto the reference value, and decode and restore the approximately complete state data within the quantization accuracy range. The decoding process does not depend on any data from previous time steps. The benchmark set update module (106) is configured to update the standardized benchmark set of the specified entity type and increment the version number in response to the benchmark set update command, and the newly collected deviation vector is associated with the new version of the benchmark set; The conservation verification module (107) is configured to perform conservation verification after decoding and restoration for dimensions with total conservation constraints and entities that are hierarchical structures containing parent and child entities. The module summarizes the state values ​​of each child entity and compares them with the total amount of the parent entity. If the deviation exceeds a preset threshold, a consistency alarm is triggered. The alarm module (108) is configured to generate an alarm event and notify the administrator when the deviation value of any dimension in the deviation vector exceeds the preset alarm threshold.

[0011] Example 7: A More Specific Encoding Example Suppose the state data of a smart meter at time t1 is as follows: Voltage: 222.5V (reference value 220.0V, deviation +2.5V) Current: 4.8A (reference value 5.0A, deviation -0.2A) Power: 1060W (baseline 1100W, deviation -40W) Power factor: 0.94 (reference value 0.95, deviation -0.01) Cumulative electricity consumption: 10025.6kWh (baseline value 10000kWh, deviation +25.6kWh) Temperature: 26.5℃ (reference value 25.0℃, deviation +1.5℃) Using the method of this invention, the deviation quantization encoding process is as follows: The voltage deviation is +2.5V, using 12-bit encoding, with an encoding range of -10V to +10V, a quantization step size of ≈0.0049V, and an encoding value of 512 (corresponding to +2.5V). The current deviation is -0.2A, using 10-bit encoding, with an encoding range of -2A to +2A, a quantization step size of ≈0.0039A, and an encoding value of -51 (corresponding to -0.2A). This process is repeated to generate a 6-dimensional offset code, which is then compressed and arranged into a 66-bit data packet. At time t2 (e.g., 15 minutes later), if the device status changes to (222.6V, 4.9A, 1080W, 0.95, 10026.0kWh, 26.0℃), the new deviation vector will be (+2.6V, -0.1A, -20W, 0, +26.0kWh, +1.0℃). Compared to time t1, the deviation vector has undergone a slight change, and the system stores the new 66-bit deviation vector. If the device state at time t3 is completely consistent with the reference set, and the deviation is zero, then a zero-value deviation vector is generated and stored using compact encoding (such as a 1-byte flag) to ensure that there is an independent record at time t3; When querying the approximately complete state of an entity at time t1, the system reads the deviation vector (66 bits) stored at time t1, locates the standardized reference set BS_0001, decodes the quantization deviation values ​​of each dimension sequentially according to the bit field layout, and superimposes them onto the reference value to restore the state: voltage ≈ 220.0 + 2.5000 = 222.5V (quantization error ≤ 0.0049V), current ≈ 5.0 - 0.1992 = 4.8008A... This allows the restoration of approximately 6-dimensional state data within the quantization accuracy. If you query the state at time t2, you read the deviation vector at time t2 and decode it. The decoding process is completely independent of the data at time t1. If you query the state at time t3, you read the zero value vector, and the result is the baseline value itself.

[0012] Example 8: Data transmission optimization in IoT device reporting scenarios Taking 10,000 smart meters as an example, each meter reports status data (6 dimensions) every 15 minutes. The traditional solution requires transmitting 192 bits (6 x 32 bits) per report, resulting in a total daily data transmission volume of 10,000 x 96 reports x 192 bits ≈ 184.3 MB ≈ 23.0 MB. Using the solution of this invention, each meter only needs to transmit 66 bits of deviation data per report, and the total daily data transmission volume is 10000 × 96 times × 66 bits ≈ 63.4M bits ≈ 7.9MB. For zero-value deviation vectors, after compact coding, a single packet can be less than 1 byte, and the savings in communication traffic are even more significant under long-term stable conditions; Overall communication traffic is reduced by approximately 65.6%. Based on NB-IoT communication tariffs, annual communication costs can be reduced by more than 65%.

[0013] Example 9: Base Set Version Upgrade Scenario When the standard operating conditions of the electricity meter change due to a software upgrade (e.g., power consumption is optimized after the upgrade, and the standard power value is adjusted from 1100W to 1050W), the administrator updates the standard status values ​​of the corresponding dimension in the standardized benchmark set, and the benchmark set version number is upgraded from V3 to V4. At this point, the newly acquired deviation vector is associated with the V4 baseline set, while the historical deviation vector remains associated with the V3 baseline set. When querying the status at a historical moment, the system automatically uses the V3 baseline set for decoding; when querying the status at the current moment, the system uses the V4 baseline set for decoding. The new and old data do not affect each other, and there is no need to migrate historical data.

[0014] (iv) Description of industrial application scenarios This invention can be widely applied to the following industrial scenarios: IoT device status monitoring: Efficiently encodes, stores, and transmits status data from a large number of sensor devices, reducing communication power consumption and storage costs. It is suitable for smart cities, smart grids, industrial IoT, and other fields. Digital twin data synchronization: In a digital twin system, the state data of the physical entity needs to be synchronized to the digital model in real time. The efficient deviation encoding scheme provided by this invention can significantly reduce the amount of data transmitted during synchronization, and improve the synchronization frequency and real-time performance; Wearable device health monitoring: Smartwatches, wristbands, and other devices continuously collect multi-dimensional health data such as heart rate, blood oxygen, sleep, and exercise. The solution of this invention can significantly reduce the amount of data transmission between the device and the mobile phone / cloud, and extend the device's battery life. Real-time financial data accounting: In collaboration with the patented "A closed-loop method for quantitative evaluation of the whole subject and object based on multidimensional value accounting and financial accounting" (A8), it provides an efficient encoding and transmission scheme for the synchronization of state data in five-dimensional value accounting, and together with the parent increment conservation scheme (specific implementation of A1), it forms a complete data link from encoding and transmission to conservation verification.

Claims

1. A method for quantizing and decoding entity state deviations based on a standardized reference set, characterized in that, include: Establish a standardized benchmark set for each entity type, which contains the standard state values ​​of the entity type in multiple dimensions; The system collects the current state data of the entity and calculates the deviation between the current state data and the standard state value of the corresponding dimension in the standardized benchmark set. The deviation value is quantized and encoded according to a preset quantization precision to generate a deviation vector. The deviation vector is associated with the entity identifier and timestamp and stored instead of storing complete state data. When the deviation value is zero in all dimensions or the change is less than a preset threshold, a zero-value deviation vector is stored to ensure that there is an independently decodeable deviation record at each sampling moment. In response to an entity state query request, the system reads the deviation vector and the standardized benchmark set, superimposes the quantized deviation value onto the benchmark value, and decodes and restores approximately complete state data within the quantization precision error range. The decoding and restoration do not depend on the state data of any previous moment; the deviation vector at any moment can be decoded independently.

2. The method according to claim 1, characterized in that, The deviation value is calculated as follows: Deviation value = Current state value - Standard state value, or Deviation value = (Current state value - Standard state value) / Standard state value.

3. The method according to claim 1, characterized in that, The quantization precision is dynamically adjusted according to the importance of the dimensions, and the quantization precision is set independently for different dimensions.

4. The method according to claim 1, characterized in that, The deviation vector is stored using bit-field compression, with the encoded data of each dimension arranged consecutively in the binary bit field according to the dimension index order.

5. The method according to claim 1, characterized in that, Also includes: When the standardized benchmark set for an entity type is updated, the system automatically updates the benchmark set version number and associates the newly collected deviation vectors with the new version of the benchmark set.

6. The method according to claim 1, characterized in that, Also includes: Maintain a time-series index of the deviation vector for each entity, and support querying historical versions of the deviation vector by time range.

7. The method according to claim 1, characterized in that, For dimensions with total quantity conservation constraints, and when the entity is divided into a hierarchical structure containing a parent entity and several child entities, a conservation verification is performed after decoding and restoration: the state values ​​of each child entity are summarized and compared with the total quantity of the parent entity. If the deviation exceeds a preset threshold, a consistency alarm is triggered.

8. The method according to claim 1, characterized in that, The zero-value deviation vector is stored using a predefined zero-value compact encoding format to further compress storage space.

9. The method according to claim 1, characterized in that, Also includes: An alarm event is generated when the deviation value of any dimension in the deviation vector exceeds a preset alarm threshold.

10. A system for quantizing and decoding entity state deviations based on a standardized reference set, characterized in that, include: The benchmark set storage module is configured to store standardized benchmark sets for each entity type. The deviation calculation module is configured to calculate the deviation between the current state data and the baseline value. The encoding module is configured to quantize and encode the deviation value to generate a deviation vector, wherein the deviation vector is stored using bit-field compression. A zero-value deviation vector is generated when the deviation is zero or less than a threshold; the storage module is configured to store the deviation vector in association with the entity identifier and timestamp, including the storage of the zero-value deviation vector, to ensure that there is an independently accessible deviation record at each sampling moment; the decoding module is configured to superimpose the deviation vector onto the reference value to restore approximately complete state data within the quantization accuracy range, and the decoding does not depend on the state data of any previous moment; The benchmark set update module is configured to update the benchmark set and increment the version number in response to the benchmark set update command.

Images