A time series data accelerated query method for a thermal power plant monitoring information system

Abstract

The application discloses a kind of time series data acceleration query method for thermal power plant monitoring information system, including step one, data acquisition and preprocessing, step two, data storage and index construction, step three, query request processing, step four, query execution and result return.The application can quickly respond to complex queries, reduce query delay, improve the real-time performance of thermal power plant monitoring system when implemented, improve data quality through data cleaning and standardization, optimize data retrieval efficiency through multi-level indexing and distributed storage, reduce system load through query optimization and parallel processing.At the same time, the scheme reduces storage space occupation and cost, and provides reliable support for the safe and stable operation of power plant.

Description

Technical Field

[0001] This invention relates to the field of thermal power plant monitoring information system technology, specifically to a method for accelerating time-series data query for thermal power plant monitoring information systems. Background Technology

[0002] Thermal power plant monitoring information systems generate massive amounts of time-series data during operation, such as parameters like temperature, pressure, and flow rate. This data is crucial for the safe operation and fault diagnosis of the power plant. However, existing time-series data query methods are often inefficient, especially when processing large-scale data, resulting in long query response times that fail to meet the demands of real-time monitoring. Traditional database systems suffer from performance bottlenecks in storing and querying time-series data, leading to data analysis delays and impacting the decision-making efficiency of operators. Furthermore, the lack of effective indexing and compression mechanisms in current technologies further exacerbates query performance issues, limiting the system's real-time performance and scalability.

[0003] Therefore, this application proposes a method for accelerating time-series data query for thermal power plant monitoring information systems. Summary of the Invention

[0004] To address this issue, the present invention provides a method for accelerating time-series data querying in thermal power plant monitoring information systems, thereby solving the problem that the lack of effective indexing and compression mechanisms in existing technologies further exacerbates query performance issues and limits the real-time performance and scalability of the system.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for accelerating time-series data query in a thermal power plant monitoring information system includes the following steps:

[0007] Step 1, Data Acquisition and Preprocessing: Real-time runtime data is collected using sensors and acquisition equipment deployed at monitoring points in the thermal power plant. The time-series data includes temperature, pressure, and flow parameters. The data is then cleaned, denoised, and standardized. Data cleaning uses a sliding window method to remove noise and interpolation to fill in missing values. Data standardization converts the data into a uniform format to eliminate outliers and improve data consistency.

[0008] Step 2, Data Storage and Index Construction: The preprocessed time-series data is stored in a time-series database, using a columnar storage format to improve compression efficiency; a multi-level index structure is constructed, including a timestamp-based B-tree index and a parameter-type-based hash index, and an index caching mechanism is established to accelerate data access;

[0009] Step 3, Query Request Processing: Receive user query requests through the query interface, parse the time range, parameter conditions, and aggregate functions in the request; rewrite and optimize the query using the index structure to generate an efficient query execution plan, including selecting the optimal index and reducing the data scan range;

[0010] Step 4, Query Execution and Result Return: Based on the query execution plan, data is retrieved from the time-series database in parallel, and in-memory computing technology is used to aggregate or calculate the results in real time. The results are then returned to the user interface in the form of charts or tables, supporting real-time streaming queries and historical data queries.

[0011] Preferably, in step one, the data cleaning also includes outlier detection and correction, specifically using statistical methods to identify outlier data that exceeds a preset threshold and replacing it with the mean or median of neighboring data; at the same time, data standardization includes normalization processing to scale the data to a uniform range.

[0012] Preferably, in step two, the time-series database adopts a distributed storage architecture, which divides the time-series data into segments according to time intervals and stores them on multiple data nodes, with each data node responsible for storing data within a specific time range;

[0013] The multi-level index structure includes a first-level global index and a second-level local index. The first-level global index is a timestamp-based distributed index used to quickly locate the node where the data is located. The second-level local index is built on each data node and includes a B+ tree index and a bitmap index based on parameter values ​​to support multi-condition queries.

[0014] During index building, index data is compressed using dictionary encoding and run-length encoding; index updates employ an incremental mechanism, periodically merging new data into the index to maintain its real-time performance; the columnar storage format supports column-by-column compression and querying, reducing I / O overhead.

[0015] Preferably, in step three, the query request processing also includes a query caching mechanism, which caches frequently used query results in memory and sets a cache expiration time; when the same or similar queries are received, the results are returned from the cache first to reduce the number of database accesses; at the same time, query optimization is based on a cost model to evaluate the cost of different indexes and scan paths and select the optimal execution plan.

[0016] Preferably, in step four, the query execution employs parallel processing technology, decomposing the query task into multiple subtasks and assigning them to different data nodes for simultaneous execution; and using a stream processing engine to continuously query real-time data; before the results are returned, the data is aggregated and calculated, including summation, average or maximum value calculation, and the results are encapsulated in JSON or XML format.

[0017] Preferably, before step two, a data compression step is also included: applying a lossless compression algorithm, including run-length encoding or dictionary encoding, to the preprocessed time-series data; and decompressing the compressed data in real time during querying.

[0018] Preferably, the data compression step is performed simultaneously with index construction, and the compressed data blocks are stored in association with the index; furthermore, during query execution, the compressed data blocks are located according to the index, and only the relevant data portions are decompressed.

[0019] Preferably, after step four, a real-time alarm step is also included: based on the query results, a threshold condition is set, and when the data exceeds the threshold, an alarm signal is triggered and the alarm information is pushed to the monitoring terminal; and an alarm log is recorded for subsequent analysis.

[0020] This invention offers the following advantages: During implementation, it enables rapid response to complex queries, reduces query latency, and improves the real-time performance of power plant monitoring systems. Data quality is enhanced through data cleaning and standardization; data retrieval efficiency is optimized through multi-level indexing and distributed storage; and system load is reduced through query optimization and parallel processing. Furthermore, this solution reduces storage space usage and costs, providing reliable support for the safe and stable operation of power plants. Attached Figure Description

[0021] To more intuitively illustrate the prior art and this application, exemplary drawings are provided below. It should be understood that the specific shapes and structures shown in the drawings should not generally be regarded as limiting conditions for implementing this application; for example, based on the technical concept disclosed in this application and the exemplary drawings, those skilled in the art are able to easily make conventional adjustments or further optimizations to the addition / reduction / classification, specific shapes, positional relationships, connection methods, size ratios, etc. of certain units (components).

[0022] Figure 1 A flowchart illustrating a method for accelerating time-series data querying in a thermal power plant monitoring information system, provided in an embodiment of this application. Detailed Implementation

[0023] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. It should be understood that these embodiments are merely for further explanation of the present invention and should not be construed as limiting the scope of protection of the present invention. Technical engineers in the field can make some non-essential improvements and adjustments to the present invention based on the above-described content. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1 A method for accelerating time-series data query in a thermal power plant monitoring information system includes the following steps:

[0025] Step 1, Data Acquisition and Preprocessing: Real-time runtime sequence data, including temperature, pressure, and flow parameters, is collected through sensors and acquisition equipment deployed at monitoring points in the thermal power plant. The data is then cleaned, denoised, and standardized. Data cleaning uses a sliding window method to remove noise and interpolation to fill in missing values. Data standardization converts the data into a uniform format to eliminate outliers and improve data consistency.

[0026] Step 2, Data Storage and Index Construction: The preprocessed time-series data is stored in a time-series database, using a columnar storage format to improve compression efficiency; a multi-level index structure is constructed, including a timestamp-based B-tree index and a parameter-type-based hash index, and an index caching mechanism is established to accelerate data access;

[0027] Step 3, Query Request Processing: Receive user query requests through the query interface, parse the time range, parameter conditions, and aggregate functions in the request; rewrite and optimize the query using the index structure to generate an efficient query execution plan, including selecting the optimal index and reducing the data scan range;

[0028] Step 4, Query Execution and Result Return: Based on the query execution plan, data is retrieved from the time-series database in parallel, and in-memory computing technology is used to aggregate or calculate the results in real time. The results are then returned to the user interface in the form of charts or tables, supporting real-time streaming queries and historical data queries.

[0029] In the implementation of this solution, in step one (data acquisition and preprocessing), the system acquires raw data in real time from sensors distributed throughout the power plant and immediately cleans, denoises, and standardizes it. This step removes invalid information and noise interference at the source, significantly improving the data quality and consistency for subsequent storage and retrieval, laying a solid foundation for high-speed and accurate queries. Subsequently, in step two (data storage and index building), the preprocessed data is stored in a database optimized for time series data, and a multi-level index is built, enabling the system to bypass full scans and instantly locate the target data block in massive amounts of data, greatly reducing I / O overhead and latency in data retrieval. When a user initiates a query, step three (query request processing) parses the user's intent and rewrites and optimizes the query using the aforementioned indexes, generating the most efficient execution plan. This avoids unnecessary waste of computing resources and intelligently improves query response speed. Finally, step four (query execution and result return), relying on the optimized plan and parallel computing technology, quickly retrieves and processes data from the storage end, presenting the results to the user in real time. The entire process is interconnected and works in synergy, meeting the stringent real-time requirements of thermal power plant monitoring. It provides operators with near-instantaneous data insights, supports their rapid decision-making, and ensures the safe and economical operation of the power plant.

[0030] In step one, the data cleaning also includes outlier detection and correction. Specifically, statistical methods are used to identify outlier data that exceeds a preset threshold and replace it with the mean or median of neighboring data. At the same time, data standardization includes normalization processing to scale the data to a uniform range in order to improve the efficiency of subsequent processing.

[0031] When implemented, the above solution effectively prevented these anomalies from misleading subsequent data analysis and trend judgment, thus improving the accuracy of monitoring and alarms. Simultaneously, the explicit normalization process ensured that parameters with different dimensions and ranges (such as temperature and pressure) could be processed and compared under the same benchmark. This not only simplified subsequent calculation logic but also made multi-parameter composite queries and model analysis possible, enhancing the standardization of system processing.

[0032] In step two, the time-series database adopts a distributed storage architecture, storing time-series data in shards across multiple data nodes according to time intervals. Each data node is responsible for storing data within a specific time range. The multi-level index structure includes a first-level global index and a second-level local index. The first-level global index is a timestamp-based distributed index used to quickly locate the node where the data is located. The second-level local index is built on each data node, including B+ tree indexes and bitmap indexes based on parameter values ​​to support multi-condition queries. During the index construction process, the index data is compressed using dictionary encoding and run-length encoding to reduce storage space and improve index loading speed. In addition, the index update adopts an incremental mechanism, periodically merging new data into the index to maintain the real-time performance of the index. The columnar storage format supports column-by-column compression and querying, reducing I / O overhead.

[0033] The introduced distributed storage architecture, which stores data in time-sharing chunks, provides excellent scalability, allowing the system to easily handle continuous data growth by adding nodes and avoiding single-machine performance bottlenecks. The first-level global index and second-level local indexes constitute a fine-grained index structure, enabling rapid narrowing of the search scope in large-scale cluster environments and precise location of specific data segments on specific nodes, greatly improving the concurrent processing capabilities of complex multi-condition queries. Furthermore, the index compression and incremental update mechanism significantly reduces storage costs and index maintenance overhead while ensuring query speed, achieving a balance between performance and resource utilization.

[0034] In step three, the query request processing also includes a query caching mechanism, which caches frequently used query results in memory and sets a cache expiration time. When the same or similar queries are received, the results are returned from the cache first to reduce the number of database accesses. At the same time, query optimization is based on a cost model to evaluate the cost of different indexes and scan paths and select the optimal execution plan.

[0035] Query caching mechanisms can temporarily store frequently accessed query results (such as the total power of the units in the most recent hour) in high-speed memory. When the same request is encountered again, the results can be returned directly, which greatly reduces the pressure of repeated accesses to the backend database and is particularly effective in improving the query speed of hot data. Meanwhile, a cost-based query optimizer, like "navigation software," can intelligently evaluate the cost of different query paths (such as whether to use index A or index B) and automatically select the optimal solution, thereby avoiding inefficient full table scans and ensuring the efficiency and stability of the query execution plan.

[0036] In step four, the query execution employs parallel processing technology, decomposing the query task into multiple subtasks and distributing them to different data nodes for simultaneous execution; and using a stream processing engine to continuously query real-time data to achieve low-latency response; before the results are returned, the data is aggregated and calculated, including summation, average or maximum value calculation, and the results are encapsulated in JSON or XML format.

[0037] By breaking down large query tasks and executing them in parallel across multiple data nodes, the advantage is that it fully utilizes the computing power of distributed systems. What was originally a serial, long-running query task is transformed into multiple short tasks running simultaneously, significantly reducing the overall response time. This is particularly suitable for statistical analysis of large historical datasets. Simultaneously, the integrated stream processing engine supports continuous querying of real-time data streams, enabling the system to continuously monitor data changes and promptly trigger events (such as limit-breaking alarms). This meets the rigid requirements for real-time situational awareness in thermal power plant monitoring, achieving a capability upgrade from "passive querying" to "proactive pushing."

[0038] Before step two, a data compression step is also included: applying lossless compression algorithms, including run-length encoding or dictionary encoding, to the preprocessed time-series data to reduce storage space usage; and decompressing the compressed data in real time during querying to balance storage efficiency and query performance. By decompressing in real time during querying, it is ensured that the improvement of storage efficiency will not come at the expense of query speed, enabling the system to store and process data with a longer time span with limited resources.

[0039] The data compression step is performed simultaneously with index building, and the compressed data blocks are stored in association with the index. Furthermore, during query execution, the compressed data blocks are located based on the index, and only the relevant data portions are decompressed to minimize processing overhead. This solution further refines the resource overhead during the query process, concentrating computational and I / O resources on processing the data portions most relevant to the query, thereby further improving query response efficiency and processing throughput on top of the already optimized approach.

[0040] Following step four, a real-time alarm step is also included: based on the query results, threshold conditions are set, and an alarm signal is triggered when the data exceeds the threshold, pushing the alarm information to the monitoring terminal; simultaneously, alarm logs are recorded for subsequent analysis. This transforms efficient data query capabilities into a direct safety assurance measure. Once a monitored parameter (such as boiler overheating) exceeds a safety threshold, the system can immediately trigger an alarm and push notifications, buying valuable emergency response time for operators. At the same time, recording alarm logs provides complete and reliable data for post-incident tracing, equipment performance analysis, and operational procedure optimization, forming a closed-loop management system from real-time monitoring to continuous improvement.

[0041] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for accelerating time-series data query in a thermal power plant monitoring information system, characterized in that, Includes the following steps: Step 1, Data Acquisition and Preprocessing: Real-time runtime data is collected using sensors and acquisition equipment deployed at monitoring points in the thermal power plant. The time-series data includes temperature, pressure, and flow parameters. The data is then cleaned, denoised, and standardized. Data cleaning uses a sliding window method to remove noise and interpolation to fill in missing values. Data standardization converts the data into a uniform format to eliminate outliers and improve data consistency. Step 2, Data Storage and Index Construction: The preprocessed time-series data is stored in a time-series database, using a columnar storage format to improve compression efficiency; a multi-level index structure is constructed, including a timestamp-based B-tree index and a parameter-type-based hash index, and an index caching mechanism is established to accelerate data access; Step 3, Query Request Processing: Receive user query requests through the query interface, parse the time range, parameter conditions, and aggregate functions in the request; rewrite and optimize the query using the index structure to generate an efficient query execution plan, including selecting the optimal index and reducing the data scan range; Step 4, Query Execution and Result Return: Based on the query execution plan, data is retrieved from the time-series database in parallel, and in-memory computing technology is used to aggregate or calculate the results in real time. The results are then returned to the user interface in the form of charts or tables, supporting real-time streaming queries and historical data queries.

2. The method for accelerating time-series data query in a thermal power plant monitoring information system according to claim 1, characterized in that, In step one, the data cleaning also includes outlier detection and correction. Specifically, it uses statistical methods to identify outlier data that exceeds a preset threshold and replaces it with the mean or median of neighboring data. At the same time, data standardization includes normalization processing to scale the data to a uniform range.

3. The method for accelerating time-series data query in a thermal power plant monitoring information system according to claim 2, characterized in that, In step two, the time-series database adopts a distributed storage architecture, which divides the time-series data into time intervals and stores them on multiple data nodes. Each data node is responsible for storing data within a specific time range. The multi-level index structure includes a first-level global index and a second-level local index. The first-level global index is a timestamp-based distributed index used to quickly locate the node where the data is located. The second-level local index is built on each data node and includes a B+ tree index and a bitmap index based on parameter values ​​to support multi-condition queries. During the index building process, the index data is compressed using dictionary encoding and run-length encoding; The index updates adopt an incremental mechanism, periodically merging new data into the index to maintain its real-time performance; the columnar storage format supports column-by-column compression and querying, reducing I / O overhead.

4. The method for accelerating time-series data query in a thermal power plant monitoring information system according to claim 3, characterized in that, In step three, the query request processing also includes a query caching mechanism, which caches frequently used query results in memory and sets a cache expiration time. When receiving the same or similar queries, results are returned from the cache first to reduce the number of database accesses; at the same time, query optimization is based on a cost model to evaluate the cost of different indexes and scan paths and select the optimal execution plan.

5. The method for accelerating time-series data query in a thermal power plant monitoring information system according to claim 4, characterized in that, In step four, the query execution employs parallel processing technology, decomposing the query task into multiple subtasks and distributing them to different data nodes for simultaneous execution; and using a stream processing engine to continuously query real-time data; before the results are returned, the data is aggregated and calculated, including summation, average or maximum value calculation, and the results are encapsulated in JSON or XML format.

6. The method for accelerating time-series data query in a thermal power plant monitoring information system according to claim 1, characterized in that, Before step two, a data compression step is also included: applying a lossless compression algorithm, including run-length encoding or dictionary encoding, to the preprocessed time-series data; and decompressing the compressed data in real time during the query.

7. The method for accelerating time-series data query in a thermal power plant monitoring information system according to claim 6, characterized in that, The data compression step is performed simultaneously with index construction, and the compressed data blocks are stored in association with the index; furthermore, during query execution, the compressed data blocks are located according to the index, and only the relevant data parts are decompressed.

8. The method for accelerating time-series data query in a thermal power plant monitoring information system according to claim 1, characterized in that, Following step four, a real-time alarm step is also included: based on the query results, a threshold condition is set, and an alarm signal is triggered when the data exceeds the threshold, and the alarm information is pushed to the monitoring terminal; and an alarm log is recorded for subsequent analysis.

Images