Automatic generation method based on characteristic value analogy report of same-type units of large hydropower station

Abstract

On the basis of the automatic generation method for the characteristic value analogy report of the same-type units of the large hydropower station, original data is collected by combining a data collecting and monitoring system, data is cleaned by using a data cleaning model based on a dynamic neural network, and the data cleaning efficiency is improved on the premise of improving the quality of the collected data. A report is intelligently generated by utilizing machine learning and a data analysis model, so that the problems of difficulty in transverse comparison of units of the same type, long time consumption and poor manual copying effect, difficulty in data extraction and processing, manpower resource consumption in report design and maintenance and the like are solved, and the aim of transverse comparison analysis visualization of operating parameters of the units of the same type is fulfilled; and the operation and maintenance management efficiency of the hydropower station equipment is greatly improved. Analysis of eigenvalue analogy data of units of the same type can clearly judge the operation state of the units, predict risks and find faults in time, and the method is of great significance to safety and economical efficiency of unit operation.

Description

Technical Field

[0001] The invention belongs to the technical field of automatic report generation, and in particular relates to an automatic generation method of analogy reports based on characteristic values ​​of similar units of large hydropower stations. Background Art

[0002] With the continuous development and progress of information technology, the analysis and utilization of big data has become the key to the development of various industries, and the power industry is no exception. When there are multiple types of units in a large hydropower station, the multiple types of units lead to different fixed value data, which brings difficulties to the operation and maintenance management of the units. By analyzing the analog data of the characteristic values ​​of the same type of units, it is possible to make a clear judgment on the operating status of the unit, predict risks, and discover faults in time, which is of great significance to the safety and economy of the unit operation. At present, the production of common analog reports of characteristic values ​​of the same type of units mostly relies on manual data extraction, extracting the required data information from the collected raw data, manually copying it into the worksheet at a fixed time and point, and then manually analyzing and judging it.

[0003] Disadvantages of existing technology: (1) Manual report entry has the following shortcomings: the data is not recorded consistently, the process takes a long time, the results are not ideal, special data is easily missed, and report design and maintenance consume human resources.

[0004] (2) The data analysis of the report is based on the collected original database. However, the original data collected by the sensor contains missing data, duplicate data, abnormal data, etc., and missed readings and multiple readings often occur. Using this method for analysis will lead to large deviations in the analysis results and affect the judgment of the operating status. (3) There are many types of units with different set values, making horizontal data comparison difficult and time-consuming. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an automatic generation method of analogy reports based on characteristic values ​​of similar units of large hydropower stations, which solves the problems of difficulty in horizontal comparison of similar units, time-consuming and poor effect of manual transcription, difficulty in multi-data extraction and processing, and human resource consumption in report design and maintenance.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: The automatic generation method of the analogy report of characteristic values ​​of the same type of units in large hydropower stations includes the following steps: Step 1: Determine the template content of the report to be generated, the template content includes equipment parameter items and corresponding characteristic values; the characteristic values ​​include maximum value, minimum value, average value and difference; Step 2: According to the equipment parameter items required in step 1, the original data is automatically extracted from the original database collected by the SCADA system; Step 3: Use a data cleaning model based on a dynamic neural network to clean the equipment parameter item data extracted in step 2. The cleaning includes detecting missing data, removing duplicate data, identifying and removing abnormal data, and interpolating values ​​to reconstruct the original data. Step 4: For the reconstructed data cleaned in step 3, the corresponding characteristic values ​​are automatically calculated according to the report requirements, and automatically filled into the report determined in step 1 according to the table template; Step 5: Use machine learning and data analysis models to extract and organize the report data generated in step 4 and generate corresponding analysis charts; Step 6: Manually review and modify the chart contents generated in Step 4 and Step 5. If they meet the requirements, output the corresponding content. If they do not meet the requirements, return to Step 1 and extract and generate the report data again. Preferably, in step 1, the equipment parameter items include: generator set model, unit serial number, unit head, unit active power, unit reactive power, unit stator slot temperature, unit stator core temperature, unit tooth pressure plate temperature, unit air cooler temperature and unit pure water temperature.

[0007] Preferably, the characteristic values ​​corresponding to the equipment parameter items include the following categories: ① The stator slot temperature of the unit includes: maximum temperature, minimum temperature, maximum and low temperature difference, and effective temperature; ② The stator core temperature of the unit includes: maximum temperature, minimum temperature, maximum low temperature difference, and effective temperature; ③ The unit gear plate temperature includes: the maximum temperature of the upper gear plate, the average temperature of the upper gear plate, the maximum temperature of the lower gear plate, and the average temperature of the lower gear plate; ④ The unit air cooler temperature includes: air cooler inlet water temperature, air cooler outlet water temperature, and air cooler cooling water flow; ⑤ The pure water temperature of the unit includes: pure water inlet temperature, pure water outlet temperature, expansion water tank water level, main circuit conductivity average value, and pure water flow average value.

[0008] Preferably, in step 2, the SCADA system is used for data collection and statistics of the working status of the unit.

[0009] Preferably, in step 3, a data cleaning model based on a dynamic neural network is used to clean the equipment parameter item data extracted in step 2, wherein the data cleaning is mainly divided into four parts: ① Detect duplicate records in the original SCADA collected data and eliminate duplicate data; ② Detect missing record data in the original SCADA collected data and complete the missing data; ③ Identify abnormal data in the original data; the abnormal data refers to data where the recorded value and the actual value are inconsistent due to various reasons; ④Correct abnormal data through data interpolation (missing data is also a kind of abnormal data and can be supplemented using the same method).

[0010] Preferably, in step 4, characteristic values ​​such as maximum value, minimum value, average value and difference value within the selected time period are calculated for the cleaned equipment parameter item data, and the calculation results are filled into the report template determined in step 1.

[0011] Preferably, in step 5, machine learning and data analysis models are used to extract and organize the generated report content, perform horizontal comparisons on parameters of similar units, obtain the range of variation of each parameter value under normal operating conditions of the units, and then generate corresponding conclusion analysis diagrams and analysis tables.

[0012] Preferably, in step 6, the tables and graphs generated in steps 4 and 5 are manually reviewed and corrected. If there are errors in the report or conclusion, steps 1 to 5 are repeated until they are confirmed to be correct, and finally a complete report that meets the requirements is output.

[0013] A system for automatically generating analogy reports based on characteristic values ​​of similar units in large hydropower stations comprises a processing module and a storage module. The program instructions executed by the processing module are the method for automatically generating analogy reports based on characteristic values ​​of similar units in large hydropower stations.

[0014] A processing module, wherein the program instructions executed by the processing module are the automatic generation method of analogy reports based on characteristic values ​​of similar units of large hydropower stations.

[0015] The present invention can achieve the following beneficial effects: The present invention mainly aims at the problems that the horizontal comparison of multi-state parameters of current large-scale hydropower station units is difficult, the manual transcription of complex data is prone to errors and omissions of important points, the efficiency is poor, and the reports cannot meet the analysis needs. According to the actual situation, the present invention proposes the combined application of process control and automatic report generation method to realize intelligent report generation, which greatly reduces manual operations and improves the consistency and accuracy of report data, laying a reliable data foundation for report analysis, risk prediction, and fault anticipation, which is of great significance to the safe and stable operation of hydropower station units. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 Automatically generate a flow chart for the analog report of characteristic values ​​of the same type of units of the present invention; Figure 2 This is the SCADA system architecture diagram of the present invention; Figure 3 This is a flow chart of the data cleaning process in step 3 of the present invention; Figure 4 This is a layout diagram of the analog analysis module for characteristic values ​​of the same type of unit in the present invention; Figure 5 This is a temperature analogy analysis report of the VGS and the generator of the Dongfeng Motor Group of the present invention. DETAILED DESCRIPTION

[0017] The preferred solution is Figures 1 to 5 As shown, the present invention proposes an automatic generation method for analogy reports of characteristic values ​​of similar units in large hydropower stations. The method combines the data collected by the Supervisory Control and Data Acquisition (SCADA) system and uses a data cleaning model based on a dynamic neural network to clean the data. On the premise of improving the quality of collected data, machine learning and data analysis models are used to intelligently generate reports, which solves the problems of difficulty in horizontal comparison of similar units, time-consuming and ineffective manual transcription, difficulty in multiple data extraction and processing, and human resource consumption in report design and maintenance. The method achieves the purpose of visualizing the horizontal comparison and analysis of the operating parameters of similar units, and greatly improves the operation and maintenance management efficiency of hydropower station equipment. The flow chart is shown in the attached figure. Figure 1 As shown, the specific steps are as follows: Step 1: Determine the template content of the report to be generated: equipment parameter items and corresponding characteristic values ​​(maximum value, minimum value, average value, etc.); Step 2: According to the equipment parameter items required in step 1, the original data is automatically extracted from the original database collected by the SCADA system; Step 3: Use a data cleaning model based on a dynamic neural network to clean the equipment parameter item data extracted in step 2. The cleaning includes detecting missing data, removing duplicate data, identifying and removing abnormal data, and interpolating values ​​to reconstruct the original data. Step 4: For the reconstructed data cleaned in step 3, the corresponding characteristic values ​​(maximum value, minimum value, average value, difference, etc.) are automatically calculated according to the report requirements, and automatically filled into the report determined in step 1 according to the table template.

[0018] Step 5: Use machine learning and data analysis models to extract and organize the report data generated in step 4 and generate corresponding analysis charts.

[0019] Step 6: Manually review and modify the chart contents generated in Step 4 and Step 5. If they meet the requirements, output the corresponding content. If they do not meet the requirements, return to Step 1 and extract and generate the report data again.

[0020] In step 1, the template format of the report content to be generated is determined. Due to the length limitation of the present invention, taking the water-cooled turbine generator temperature analogy analysis report as an example, the report content design in this step specifically includes the following equipment parameter items: generator set model, unit serial number, unit head, unit active power, unit reactive power, unit stator slot temperature, unit stator core temperature, unit tooth pressure plate temperature, unit air cooler temperature, unit pure water temperature, etc. Among them, the characteristic values ​​corresponding to the equipment parameter items are respectively: ① The stator slot temperature of the unit includes: maximum temperature, minimum temperature, maximum and low temperature difference, and effective temperature; ② The stator core temperature of the unit includes: maximum temperature, minimum temperature, maximum low temperature difference, and effective temperature; ③ The unit gear plate temperature includes: the maximum temperature of the upper gear plate, the average temperature of the upper gear plate, the maximum temperature of the lower gear plate, and the average temperature of the lower gear plate; ④ The unit air cooler temperature includes: air cooler inlet water temperature, air cooler outlet water temperature, and air cooler cooling water flow; ⑤ The pure water temperature of the unit includes: pure water inlet temperature, pure water outlet temperature, expansion water tank water level, main circuit conductivity average value, and pure water flow average value.

[0021] In step 2, the equipment parameter item data required in step 1 are automatically extracted from the original database collected by the SCADA system. The SCADA system is mainly used for data collection and statistics of the unit working status. The SCADA system architecture is shown in the attached Figure 2 .

[0022] In step 3, the data cleaning model based on dynamic neural network is used to clean the equipment parameter item data extracted in step 2. The data cleaning flow chart is as follows (see attached Figure 3 ). Data cleaning is mainly divided into four parts: ① Detect duplicate records in the original SCADA collected data and eliminate duplicate data; ② Detect missing record data in the original SCADA collected data and complete the missing data; ③ Identify abnormal data in the original data (i.e. data where the recorded value does not match the actual value due to various reasons); ④Correct abnormal data through data interpolation (missing data is also a kind of abnormal data and can be supplemented using the same method).

[0023] In step 4, characteristic values ​​such as maximum value, minimum value, average value and difference value within the selected time period are calculated for the cleaned equipment parameter item data, and the calculation results are filled into the report template determined in step 1.

[0024] In step 5, machine learning and data analysis models are used to extract and organize the generated report content, perform horizontal comparisons on parameters of similar units, obtain the range of variation of each parameter value under normal operating conditions of the units, and then generate corresponding conclusion analysis diagrams and analysis tables.

[0025] In step 6, the tables and graphs generated in steps 4 and 5 are manually reviewed and corrected. If there are errors in the report or conclusion, steps 1 to 5 are repeated until they are confirmed to be correct, and finally a complete report that meets the requirements is output.

[0026] The present invention realizes the automatic generation of characteristic value analogy reports of the same type of units in a large hydropower station. The report analysis module is mainly divided into four parts (see attached Figure 4 ): Analogy analysis of three bearings of ALSTOM and Harbin Electric Power Unit, Analogy analysis of generator temperature of ALSTOM and Harbin Electric Power Unit, Analogy analysis of generator temperature of VGS and Dongfang Electric Power Unit, and Analogy analysis of three bearings of VGS and Dongfang Electric Power Unit. Taking the analogy of generator temperature of VGS and Dongfang Electric Power Unit as an example, the method of generating a horizontal comparison report of the data of the same type of units is demonstrated: Select the VGS and Dongfang Electric Power Unit generator temperature module in the function module, select the time period to be analyzed, and click the query button to generate the corresponding analysis report. This process executes steps 1 to 5; finally, execute step 6 for manual review and output complete report content that meets the requirements. See the attached for specific example output report Figure 5 .

[0027] The present invention independently designs a report content template required for parameter comparison of the same type of units: equipment parameter items and corresponding characteristic values; combines the SCADA system for data acquisition, extracts original data from the SCADA system database, and combines the data cleaning model based on the dynamic neural network to clean and reconstruct the extracted equipment parameter item data; combines machine learning and data analysis models to intelligently generate report content, obtain the changing rules within the data, and obtain conclusion analysis charts.

[0028] The present invention greatly improves the quality of report data, solves the problems of difficulty in horizontal comparison of similar units, time-consuming and ineffective manual transcription, difficulty in multiple data extraction and processing, and human resource consumption in report design and maintenance, and achieves the purpose of visualizing horizontal comparative analysis of operating parameters of similar units, greatly improves the operation and maintenance management efficiency of hydropower station equipment, and increases the safety and reliability of unit operation.

[0029] The above embodiments are only preferred technical solutions of the present invention and should not be regarded as limiting the present invention. The protection scope of the present invention shall be the technical solutions recorded in the claims, including equivalent replacement solutions of the technical features in the technical solutions recorded in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present invention.

Claims

1. The automatic generation method of analogy report of characteristic values ​​of similar units in large hydropower stations is characterized by The following steps are involved: Step 1: Determine the template content of the report to be generated, the template content includes equipment parameter items and corresponding characteristic values; the characteristic values ​​include maximum value, minimum value, average value and difference; Step 2: According to the equipment parameter items required in step 1, the original data is automatically extracted from the original database collected by the SCADA system; Step 3: Use a data cleaning model based on a dynamic neural network to clean the equipment parameter item data extracted in step 2. The cleaning includes detecting missing data, removing duplicate data, identifying and removing abnormal data, and interpolating values ​​to reconstruct the original data. Step 4: For the reconstructed data cleaned in step 3, the corresponding characteristic values ​​are automatically calculated according to the report requirements, and automatically filled into the report determined in step 1 according to the table template; Step 5: Use machine learning and data analysis models to extract and organize the report data generated in step 4 and generate corresponding analysis charts; Step 6: Manually review and modify the chart contents generated in Step 4 and Step 5. If they meet the requirements, output the corresponding content. If they do not meet the requirements, return to Step 1 and extract and generate the report data again.

2. The method for automatically generating a characteristic value analogy report based on similar units of a large hydropower station according to claim 1 is characterized in that: In step 1, the equipment parameter items include: generator set model, unit serial number, unit head, unit active power, unit reactive power, unit stator slot temperature, unit stator core temperature, unit tooth pressure plate temperature, unit air cooler temperature and unit pure water temperature.

3. The method for automatically generating a characteristic value analogy report based on similar units of large hydropower stations according to claim 2 is characterized in that: The characteristic values ​​corresponding to the device parameter items include the following categories: ① The stator slot temperature of the unit includes: maximum temperature, minimum temperature, maximum and low temperature difference, and effective temperature; ② The stator core temperature of the unit includes: maximum temperature, minimum temperature, maximum low temperature difference, and effective temperature; ③ The unit gear plate temperature includes: the maximum temperature of the upper gear plate, the average temperature of the upper gear plate, the maximum temperature of the lower gear plate, and the average temperature of the lower gear plate; ④ The unit air cooler temperature includes: air cooler inlet water temperature, air cooler outlet water temperature, and air cooler cooling water flow; ⑤ The pure water temperature of the unit includes: pure water inlet temperature, pure water outlet temperature, expansion water tank water level, main circuit conductivity average value, and pure water flow average value.

4. The method for automatically generating a characteristic value analogy report based on similar units of a large hydropower station according to claim 1, characterized in that: In step 2, the SCADA system is used for data collection and statistics of the unit's operating status.

5. The method for automatically generating a characteristic value analogy report based on similar units of a large hydropower station according to claim 1, characterized in that: In step 3, a data cleaning model based on a dynamic neural network is used to clean the equipment parameter item data extracted in step 2, wherein the data cleaning is mainly divided into four parts: ① Detect duplicate records in the original SCADA collected data and eliminate duplicate data; ② Detect missing record data in the original SCADA collected data and complete the missing data; ③Identify abnormal data in the original data; The abnormal data refers to data where the recorded value and the actual value do not match due to various reasons; ④Correct abnormal data through data interpolation.

6. The method for automatically generating a characteristic value analogy report based on similar units of a large hydropower station according to claim 1, characterized in that: In step 4, characteristic values ​​such as maximum value, minimum value, average value and difference value within the selected time period are calculated for the cleaned equipment parameter item data, and the calculation results are filled into the report template determined in step 1.

7. The method for automatically generating a characteristic value analogy report based on similar units of a large hydropower station according to claim 1, characterized in that: In step 5, machine learning and data analysis models are used to extract and organize the generated report content, perform horizontal comparisons on parameters of similar units, obtain the range of variation of each parameter value under normal operating conditions of the units, and then generate corresponding conclusion analysis diagrams and analysis tables.

8. The method for automatically generating a characteristic value analogy report based on similar units of a large hydropower station according to claim 1 is characterized in that: In step 6, the tables and graphs generated in steps 4 and 5 are manually reviewed and corrected. If there are errors in the report or conclusion, steps 1 to 5 are repeated until they are confirmed to be correct, and finally a complete report that meets the requirements is output.

9. An automatic generation system based on analogy reports of characteristic values ​​of similar units of large hydropower stations, comprising a processing module and a storage module, characterized in that: The program instructions executed by the processing module are the automatic generation method of the analogy report based on the characteristic values ​​of the same type of units in a large hydropower station according to any one of claims 1 to 8.

10. A processing module, characterized in that: The program instructions executed by the processing module are the automatic generation method of the analogy report based on the characteristic values ​​of the same type of units in a large hydropower station according to any one of claims 1 to 8.

Images