A generator air cooler fault early warning method and system and a storage medium

Abstract

The application provides a generator air cooler fault early warning method and system and a storage medium. The method comprises the following steps: collecting unit monitoring data of multiple air coolers in a target generator unit through a unit monitoring system of the target generator unit; preprocessing external parameter data into parameter assignment data; performing data cleaning and data padding processing on the unit monitoring data; calculating and processing all the unit monitoring data based on preset temperature algorithm parameters and in combination with the state time series data to obtain temperature characteristic values of the multiple air coolers; generating a temperature characteristic index in combination with the parameter assignment data and the temperature characteristic values; and generating fault early warning information corresponding to the air coolers in combination with a preset early warning threshold and the temperature characteristic index. The application has the effect of automatically warning the signs of air cooler faults of hydroelectric generator units.

Description

Technical Field

[0001] This invention belongs to the field of air cooler fault detection technology, specifically relating to a generator air cooler fault early warning method, system, and storage medium. Background Technology

[0002] As a crucial component of hydroelectric generator sets, the generator experiences gradual temperature increases in its core and coils due to copper and iron losses during operation. To reduce the aging rate of the coil insulation and extend its lifespan, the generator temperature must be maintained within a reasonable range. Therefore, the generator air cooling system plays a vital auxiliary role. Typically, the generator air cooling system uses air as the cooling medium to cool the generator stator and rotor windings, as well as the stator core. The air cooler, a key component of the cooling system, uses cooling water to cool the high-temperature air exiting the generator before it is reintroduced. Currently, hydroelectric plants typically monitor the simulated inlet (hot air) and outlet (cold air) temperatures of the generator air cooler to indirectly assess its normal operating status and cooling effect, relying on manual experience to determine if faults such as air cooler blockage have occurred. Based on the inlet (hot air) temperature and outlet (cold air) temperature of the air cooler, characteristic indicators such as temperature calculation quantity, temperature range, hot and cold air temperature difference, and hot and cold air temperature difference dispersion can be calculated. These indicators can provide symptom indicators for judging faults such as decreased air cooler efficiency and air cooler blockage, help accurately locate air cooler faults, and provide support for condition-based maintenance of power generation air coolers.

[0003] However, the aforementioned air cooler condition assessment methods have the following shortcomings: Conventional hydropower unit air cooler temperature indicators are limited to the inlet and outlet temperatures, lacking further data mining and feature extraction, thus failing to deeply explore the value of big data. Furthermore, conventional hydropower units rely heavily on manual experience to judge air cooler efficiency decline, which is subjective and cannot be based on the long-term performance of air cooler temperature characteristic indicators, thus failing to automatically provide early warnings of fault symptoms. Summary of the Invention

[0004] This invention provides a method, system, and storage medium for early warning of generator air cooler faults, in order to solve the problem of difficulty in achieving automatic early warning of fault symptoms in hydropower unit air coolers.

[0005] In a first aspect, the present invention provides a method for early warning of generator air cooler faults, the method comprising the following steps:

[0006] The unit monitoring system of the target generator set collects unit monitoring data of multiple air coolers in the target generator set. The unit monitoring data includes unit measurement point data, status timing data and external parameter data.

[0007] The external parameter data is preprocessed into parameter assignment data;

[0008] The data from the unit's measuring points were cleaned and supplemented.

[0009] Based on preset temperature algorithm parameters and combined with the state time series data, all the unit measurement point data are calculated and processed to obtain multiple temperature characteristic values ​​of the air cooler;

[0010] Temperature characteristic indices are generated by combining the formal parameter assignment data and the temperature characteristic values;

[0011] The system combines the preset warning threshold and the temperature characteristic indicators to generate fault warning information corresponding to the air cooler.

[0012] Optionally, the preset temperature algorithm parameters include inlet and outlet temperature algorithm parameters and hot and cold air temperature algorithm parameters. The temperature feature values ​​include the calculated inlet temperature, calculated outlet temperature, inlet temperature range, outlet temperature range, hot and cold air temperature difference, and hot and cold air temperature difference dispersion of the air cooler.

[0013] Optionally, the step of calculating and processing all the unit's measuring point data based on preset temperature algorithm parameters and combined with the state time series data to obtain multiple temperature characteristic values ​​of the air cooler includes the following steps:

[0014] According to the generator set number corresponding to the unit measurement point data, all the unit measurement point data are grouped into a first measurement point data group and a second measurement point data group.

[0015] The unit measurement point data in the first measurement point data group and the second measurement point data group are traversed sequentially, and the unit measurement point data in each measurement point data group are sorted according to the measurement point data time axis;

[0016] Based on the state time series data, data filtering is performed on the sorted first measurement point data group and the second measurement point data group respectively;

[0017] The inlet and outlet temperature algorithm parameters are used to calculate and process the unit measuring point data in the first measuring point data group to obtain the inlet and outlet temperature calculation quantity and the inlet and outlet temperature range.

[0018] The unit measurement point data in the second measurement point data group are calculated and processed using the cold and hot air temperature algorithm parameters to obtain the cold and hot air temperature difference and the dispersion of the cold and hot air temperature difference.

[0019] Optionally, the status time series data includes unit start-up duration and unit start-up delay, and the unit measurement point data includes unit active power, unit start-up status, and unit start-up time. The step of filtering data based on the status time series data for the sorted first measurement point data group and the second measurement point data group includes the following steps:

[0020] Filter out all unit measurement point data in the first and second measurement point data groups where the unit is in the off state;

[0021] Determine whether the unit startup time is greater than or equal to the sum of the corresponding unit startup delay and the unit startup time;

[0022] If the unit start-up time is less than the sum of the corresponding unit start-up delay and the unit start-up time, then the corresponding unit measurement point data will be filtered out.

[0023] If the unit start-up time is greater than or equal to the sum of the corresponding unit start-up delay and the unit start-up time, then it is determined whether the corresponding active power of the unit is greater than a preset power threshold.

[0024] If the active power of the corresponding unit is less than or equal to the power threshold, then the corresponding unit measurement point data is filtered out.

[0025] If the active power of the corresponding unit is greater than the power threshold, the corresponding unit measurement data is retained.

[0026] Optionally, the unit measurement data includes the air cooler inlet temperature and the air cooler outlet temperature. The step of calculating and processing the unit measurement data in the first measurement data set using the inlet and outlet temperature algorithm parameters to obtain the calculated inlet and outlet temperatures and the inlet and outlet temperature range includes the following steps:

[0027] According to the characteristic fields of the air cooler inlet temperature and the air cooler outlet temperature, all the unit measurement point data in the first measurement point data group are traversed, and all the air cooler inlet temperatures in the first measurement point data group are divided into inlet temperature group, and all the air cooler outlet temperatures in the first measurement point data group are divided into outlet temperature group.

[0028] Rearrange all data in the inlet temperature group and the outlet temperature group in descending order;

[0029] Based on the inlet and outlet temperature algorithm parameters, the average value of the first three air cooler inlet temperatures in the inlet temperature group is calculated as the inlet temperature calculation value, and the average value of the first three air cooler outlet temperatures in the outlet temperature group is calculated as the outlet temperature calculation value.

[0030] Based on the inlet and outlet temperature algorithm parameters, the difference between the maximum and minimum values ​​of the air cooler inlet temperature in the inlet temperature group is calculated as the inlet temperature range, and the difference between the maximum and minimum values ​​of the air cooler outlet temperature in the outlet temperature group is calculated as the outlet temperature range.

[0031] Optionally, the unit measurement data includes the air cooler cold air temperature and the air cooler hot air temperature. The calculation and processing of the unit measurement data in the second measurement data group using the cold and hot air temperature algorithm parameters to obtain the cold and hot air temperature difference and the dispersion of the cold and hot air temperature difference includes the following steps:

[0032] Based on each of the unit measurement point data, all the unit measurement point data in the second measurement point data group are traversed, and the air cooler cold air temperature and the air cooler hot air temperature belonging to the same measurement point in the second measurement point data group are divided into an air cooler temperature group.

[0033] Based on the cold and hot air temperature algorithm parameters, the difference between the cold air temperature and the hot air temperature of the air cooler in each of the air cooler temperature groups is calculated to obtain the cold and hot air temperature difference in discrete data form.

[0034] Calculate the data dispersion of all the hot and cold air temperature differences, and use the data dispersion as the dispersion of the hot and cold air temperature differences.

[0035] Optionally, the data cleaning and data completion processing of the unit's measuring point data includes the following steps:

[0036] Iterate through all the unit's measurement point data and filter out abnormal data in the unit's measurement point data;

[0037] Calculate the mean and standard deviation of all the unit measurement data;

[0038] All unit measurement point data that differ from the average value by more than 3 times the standard deviation are removed;

[0039] The data of the unit's measuring points after screening are supplemented in seconds according to the method of pre-complementation.

[0040] Optionally, the step of generating fault warning information corresponding to the air cooler by combining the preset warning threshold and the temperature characteristic index includes the following steps:

[0041] Compare all the temperature characteristic indicators with the preset warning thresholds one by one;

[0042] If there is a target temperature feature index that exceeds the warning threshold, then extract the index name of the target temperature feature index;

[0043] The number of the target temperature characteristic indicators and the duration of each target temperature characteristic indicator exceeding the warning threshold are counted.

[0044] By combining the number of indicators and the market exceeding the standard, a warning level corresponding to the target temperature characteristic indicator is generated;

[0045] By combining the indicator name and the warning level, a fault warning message corresponding to the air cooler is generated.

[0046] In a second aspect, the present invention also provides a generator air cooler fault early warning system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the generator air cooler fault early warning method as described in the first aspect.

[0047] Thirdly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the generator air cooler fault early warning method described in the first aspect.

[0048] The beneficial effects of this invention are:

[0049] The generator air cooler fault early warning method proposed in this invention presents a technical approach from three aspects: data source and cleaning, index calculation algorithm, and fault symptom early warning implementation. Data is collected in real time by the unit monitoring system. Integrated air cooler temperature monitoring data is obtained by integrating the data time stamp and fields. The data is then cleaned to obtain cleaned integrated air cooler temperature monitoring data. Generator air cooler temperature characteristic indicators are calculated and extracted in real time based on temperature algorithm parameters, resulting in the calculated and extracted temperature indicators for hydropower generator air cooler, providing a data foundation for generator air cooler fault symptom early warning and fault diagnosis. This method achieves the extraction of multiple temperature-related indicators from generator air coolers of multiple units, and in practical applications, automatic early warning of air cooler fault symptoms is realized based on the extracted air cooler temperature characteristic indicators. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating the generator air cooler fault early warning method of the present invention. Detailed Implementation

[0051] This invention discloses a method for early warning of generator air cooler faults.

[0052] Reference Figure 1 The generator air cooler fault early warning method specifically includes the following steps:

[0053] S101. Collect unit monitoring data of multiple air coolers in the target generator set through the unit monitoring system of the target generator set.

[0054] The unit monitoring system can collect unit monitoring data from air coolers 1-n in target generator sets 1-m. The unit monitoring data includes unit measurement point data, status timing data, and external parameter data. Taking a single target generator set as an example, referring to Table 1, the unit measurement point data includes, but is not limited to: air cooler hot air (inlet) temperature (extended to air coolers 1-n), air cooler cold air (outlet) temperature (extended to air coolers 1-n), unit active power, and generator outlet switch status.

[0055] Table 1. Unit measurement data of the target generator set

[0056]

[0057]

[0058] S102. Preprocess the external parameter data into parameter assignment data.

[0059] The external parameter data includes, but is not limited to: common fields for input measurement point IDs, common fields for output measurement point IDs, active power thresholds, and start-up delay time. The strings containing common fields for input measurement point IDs, common fields for output measurement point IDs, active power measurement points of the unit, and corresponding thresholds in the external parameter data are separated by commas and stored as new strings as parameter assignment data. This facilitates subsequent indexing and combination for assigning values ​​to output measurement point IDs.

[0060] S103. Perform data cleaning and data completion processing on the unit's measuring point data.

[0061] Due to factors such as line interference and loose sensors, abnormal data exists in the data, necessitating data cleaning. The cleaned data then needs to be supplemented, specifically using a pre-compensation method. Considering the differences in data acquisition frequency between operating parameters and temperature measurement points, or data loss due to data link issues, data supplementation for all measurement points is performed on a second-by-second basis using a pre-compensation method.

[0062] S104. Based on the preset temperature algorithm parameters and combined with the status time series data, calculate and process the measurement data of all units to obtain the temperature characteristic values ​​of multiple air coolers.

[0063] Referring to Table 2, the algorithm parameters are the set of important configurable parameters of the algorithm. To increase the flexibility of the algorithm, ensure configurability, and facilitate better promotion and application, some parameters of the algorithm are adjusted based on data conditions and computational results, and the important parameters of the algorithm are condensed into a set of configurable parameters.

[0064]

[0065] S105. Combine the formal parameter assignment data and temperature characteristic values ​​to generate temperature characteristic indices.

[0066] S106. Generate corresponding air cooler fault warning information by combining preset warning thresholds and temperature characteristic indicators.

[0067] The implementation principle of this method is as follows:

[0068] The generator air cooler fault early warning method proposed in this invention presents a technical approach from three aspects: data source and cleaning, index calculation algorithm, and fault symptom early warning implementation. Data is collected in real time by the unit monitoring system. Integrated air cooler temperature monitoring data is obtained by integrating the data time stamp and fields. The data is then cleaned to obtain cleaned integrated air cooler temperature monitoring data. Generator air cooler temperature characteristic indicators are calculated and extracted in real time based on temperature algorithm parameters, resulting in the calculated and extracted temperature indicators for hydropower generator air cooler, providing a data foundation for generator air cooler fault symptom early warning and fault diagnosis. This method achieves the extraction of multiple temperature-related indicators from generator air coolers of multiple units, and in practical applications, automatic early warning of air cooler fault symptoms is realized based on the extracted air cooler temperature characteristic indicators.

[0069] In one embodiment, the preset temperature algorithm parameters include inlet and outlet temperature algorithm parameters and hot and cold air temperature algorithm parameters. The temperature characteristic values ​​include the calculated amount of inlet temperature of the air cooler, the calculated amount of outlet temperature, the inlet temperature range, the outlet temperature range, the hot and cold air temperature difference, and the dispersion of the hot and cold air temperature difference.

[0070] In this embodiment, step S104 specifically includes the following steps:

[0071] According to the generator unit number corresponding to the unit's measurement point data, all unit measurement point data are grouped into the first measurement point data group and the second measurement point data group.

[0072] The unit measurement data in the first and second measurement data groups are traversed sequentially, and the unit measurement data in each measurement data group are sorted according to the measurement data time axis.

[0073] Based on the state time series data, data filtering is performed on the sorted first and second measurement point data groups respectively;

[0074] The unit measuring point data in the first measuring point data group are calculated and processed using the inlet and outlet temperature algorithm parameters to obtain the inlet and outlet temperature calculation amount and the inlet and outlet temperature range.

[0075] The unit measurement point data in the second measurement point data group are calculated and processed using the cold and hot air temperature algorithm parameters to obtain the cold and hot air temperature difference and the dispersion of the cold and hot air temperature difference.

[0076] In this embodiment, the generator set measurement data is grouped according to the generator set number to facilitate subsequent processing of data from different generator sets. Assume there are two generator sets, numbered A and B. Based on the generator set number in the measurement data, the data is divided into two groups: one containing all data for generator set number A, and the other containing all data for generator set number B. The data within each generator set measurement data group is sorted according to a time axis to ensure the accuracy and consistency of subsequent data processing. For the measurement data of a specific generator set in the first measurement data group, it is sorted according to the time sequence of the measurement data to ensure that the data is processed chronologically. Based on the characteristics of state-time series data, the sorted measurement data groups are filtered to remove abnormal or invalid data to ensure the accuracy of subsequent data processing. For example, for the measurement data of a specific generator set in the first measurement data group, abnormal or invalid data, such as values ​​outside the range or missing data, are filtered out according to the rules of state-time series data.

[0077] Based on the inlet and outlet temperature algorithm parameters, the unit measuring point data in the first measuring point data group are processed to obtain the calculated amount and range of the inlet and outlet temperatures, which are used for subsequent analysis and judgment. Then, based on the cold and hot air temperature algorithm parameters, the unit measuring point data in the second measuring point data group are processed to obtain the difference and dispersion of the cold and hot air temperatures, which are used for subsequent analysis and judgment.

[0078] In one embodiment, the state timing data includes the unit start-up duration and the unit start-up delay, and the unit measurement point data includes the unit active power, the unit start-up status, and the unit start-up time. The step of filtering data from the sorted first and second measurement point data groups based on the state timing data specifically includes the following steps:

[0079] Filter out all unit measurement point data in the first and second measurement point data groups where the unit is in the off state and the unit is in the on state.

[0080] Determine whether the unit startup time is greater than or equal to the sum of the corresponding unit startup delay and the unit startup time;

[0081] If the unit start-up time is less than the sum of the corresponding unit start-up delay and the unit start-up time, then the corresponding unit measurement point data will be filtered out.

[0082] If the unit start-up time is greater than or equal to the sum of the corresponding unit start-up delay and the unit start-up time, then it is determined whether the active power of the corresponding unit is greater than the preset power threshold.

[0083] If the active power of the corresponding unit is less than or equal to the power threshold, the corresponding unit measurement point data will be filtered out.

[0084] If the active power of the corresponding unit is greater than the power threshold, the corresponding unit measurement data will be retained.

[0085] In this implementation, based on the unit's startup status, all measurement data from units whose startup status is "shutdown" are filtered out to ensure that all data processed subsequently is from units in startup status. Then, based on the relationship between unit startup duration, startup delay, and startup time, it is determined whether the unit meets the startup duration requirement. Assume that unit B's startup duration is 10 hours, startup delay is 2 hours, and startup time is 8 hours. Determine whether unit B's startup duration is greater than or equal to 2 hours + 8 hours = 10 hours.

[0086] If the unit's start-up duration does not meet the requirement of the sum of the start-up delay and the start-up time, it indicates that the data from that unit is unreliable and needs to be filtered out. Assume that unit B's start-up duration is 9 hours, the start-up delay is 2 hours, and the start-up time is 8 hours. Since 9 hours < 2 hours + 8 hours = 10 hours, all measurement data from unit B must be filtered out. By setting the unit's start-up delay, data within t hours of start-up can be filtered out, where t is the unit's start-up delay.

[0087] If the unit's start-up duration meets the requirement of the sum of start-up delay and start-up time, then it is further determined whether the unit's active power meets the preset threshold requirement. If the unit's active power does not meet the preset threshold requirement, it indicates that the unit's data is unreliable and needs to be filtered out. Assume that unit B's active power is 80MW, and the preset power threshold is 100MW. Since 80MW <= 100MW, all measurement data for unit B are filtered out. Assume that unit B's active power is 120MW, and the preset power threshold is 100MW. Since 120MW > 100MW, all measurement data for unit B are retained.

[0088] In one embodiment, the unit measurement data includes the air cooler inlet temperature and the air cooler outlet temperature. The steps of calculating and processing the unit measurement data in the first measurement data group using inlet and outlet temperature algorithm parameters to obtain the calculated inlet and outlet temperatures and the inlet and outlet temperature range specifically include the following steps:

[0089] Based on the characteristic fields of air cooler inlet temperature and air cooler outlet temperature, traverse all machine group measurement point data in the first measurement point data group, and divide all air cooler inlet temperatures in the first measurement point data group into inlet temperature group and all air cooler outlet temperatures in the first measurement point data group into outlet temperature group.

[0090] Rearrange all data in the inlet temperature group and the outlet temperature group in descending order;

[0091] The average inlet temperature of the first three air coolers in the inlet temperature group is calculated based on the inlet and outlet temperature algorithm parameters as the inlet temperature calculation quantity, and the average outlet temperature of the first three air coolers in the outlet temperature group is calculated as the outlet temperature calculation quantity.

[0092] The difference between the maximum and minimum inlet temperatures of the air cooler in the inlet temperature group is calculated based on the inlet and outlet temperature algorithm parameters as the inlet temperature range, and the difference between the maximum and minimum outlet temperatures of the air cooler in the outlet temperature group is calculated as the outlet temperature range.

[0093] In this embodiment, based on the feature fields, the unit measurement point data in the first measurement point data group are grouped according to the air cooler inlet temperature and outlet temperature for subsequent data processing. The data in the inlet temperature group and outlet temperature group are sorted for subsequent calculations. Based on the inlet and outlet temperature algorithm parameters, the top three air cooler inlet temperatures are selected from the inlet temperature group, and their average value is calculated as the inlet temperature calculation. Similarly, the top three air cooler outlet temperatures are selected from the outlet temperature group, and their average value is calculated as the outlet temperature calculation. Assuming there are 10 air cooler inlet temperature data points in the inlet temperature group, and the top three inlet temperatures are 25℃, 23℃, and 22℃, the inlet temperature calculation is (25+23+22) / 3 = 23.33℃. Similarly, assuming there are 10 air cooler outlet temperature data points in the outlet temperature group, and the top three outlet temperatures are 30℃, 28℃, and 27℃, the outlet temperature calculation is (30+28+27) / 3 = 28.33℃.

[0094] Based on the inlet and outlet temperature algorithm parameters, the maximum and minimum inlet temperatures of the air coolers are selected from the inlet temperature group, and their difference is calculated as the inlet temperature range. Similarly, the maximum and minimum outlet temperatures of the air coolers are selected from the outlet temperature group, and their difference is calculated as the outlet temperature range. Assuming there are inlet temperature data points for 10 air coolers in the inlet temperature group, with a maximum value of 30℃ and a minimum value of 20℃, the inlet temperature range is 30℃ - 20℃ = 10℃. Likewise, assuming there are outlet temperature data points for 10 air coolers in the outlet temperature group, with a maximum value of 35℃ and a minimum value of 25℃, the outlet temperature range is 35℃ - 25℃ = 10℃.

[0095] In one embodiment, the unit measurement data includes the air-cooled air temperature and the air-cooled air temperature. The step of using the subcooled and hot air temperature algorithm parameters to calculate and process the unit measurement data in the second measurement data set to obtain the cold and hot air temperature difference and the dispersion of the cold and hot air temperature difference specifically includes the following steps:

[0096] Based on the unit measurement data, the unit measurement data is traversed through all the unit measurement data in the second measurement data group, and the air cooler cold air temperature and air cooler hot air temperature belonging to the same measurement point in the second measurement data group are divided into an air cooler temperature group.

[0097] Based on the cold and hot air temperature algorithm parameters, the difference between the cold air temperature and the hot air temperature of the air cooler in each air cooler temperature group is calculated to obtain the cold and hot air temperature difference in discrete data form.

[0098] Calculate the data dispersion of all hot and cold air temperature differences, and use the data dispersion as the dispersion of hot and cold air temperature differences.

[0099] In this embodiment, based on the various measurement points of the unit's measurement data, the air-cooled cold air temperature and air-cooled hot air temperature belonging to the same measurement point in the second measurement point data group are grouped for subsequent data processing. According to the cold and hot air temperature algorithm parameters, the difference between the air-cooled cold air temperature and the air-cooled hot air temperature in each air-cooled temperature group is calculated to obtain the cold and hot air temperature difference in discrete data form.

[0100] Suppose an air cooler temperature group contains 5 air coolers with cold air temperatures of 20℃, 22℃, 21℃, 23℃, and 20℃, and hot air temperatures of 40℃, 42℃, 41℃, 43℃, and 40℃. Based on the cold and hot air temperature algorithm parameters, the cold and hot air temperature difference for each air cooler is calculated, resulting in discrete cold and hot air temperature difference data: 2℃, 1℃, 3℃, 0℃, and 2℃. Then, based on the dispersion of the cold and hot air temperature difference data, algorithms such as the mean absolute deviation algorithm, the coefficient of variation algorithm, or the variance and standard deviation algorithm are used to calculate the dispersion of all cold and hot air temperature difference data, thus obtaining the dispersion of the cold and hot air temperature difference.

[0101] In one embodiment, step S103 specifically includes the following steps:

[0102] Iterate through all unit measurement point data and filter out abnormal data in the unit measurement point data;

[0103] Calculate the mean and standard deviation of all unit measurement data;

[0104] All unit measurement data that differ from the mean by more than three times the standard deviation were removed;

[0105] Based on the pre-complementation method, the data of the unit measurement points after screening are supplemented in seconds.

[0106] In this implementation, for each unit's measurement data, an outlier range is set or statistical methods are used to detect whether the data exceeds the normal range; if it does, it is filtered out. Statistical analysis is performed on all unit measurement data to calculate its mean and standard deviation. Specifically, the sum of all unit measurement data is calculated, then divided by the number of data points to obtain the mean. The square root of the sum of the squares of the differences between each data point and the mean is then calculated to obtain the standard deviation. Based on the mean and standard deviation, a range is determined, and data differing from the mean by more than three times the standard deviation are filtered out. For the unit measurement data after filtering out outliers, data is supplemented using a pre-complementation method, with the time precision measured in seconds. For example, if a unit's measurement data is missing at a certain time point, it is supplemented using the most recent valid data before that time point. The supplementation time precision is in the second range, meaning that supplementation is performed every second before the missing time point.

[0107] In one embodiment, step S106 specifically includes the following steps:

[0108] Compare all temperature characteristic indicators with the preset warning thresholds one by one;

[0109] If there are target temperature feature indicators that exceed the warning threshold, extract the indicator name of the target temperature feature indicator;

[0110] The number of target temperature characteristic indicators and the duration of each target temperature characteristic indicator exceeding the warning threshold are recorded.

[0111] The warning level is generated by combining the number of indicators and the market exceeding the standard, corresponding target temperature characteristic indicators.

[0112] By combining the indicator name and the warning level, corresponding fault warning information for the air cooler is generated.

[0113] In this embodiment, each temperature characteristic indicator is compared with a preset warning threshold to determine whether it exceeds the threshold. If a target temperature characteristic indicator exceeds the warning threshold, its name is extracted. The number of target temperature characteristic indicators exceeding the warning threshold and the duration of each indicator exceeding the threshold are counted. For example, if three target temperature characteristic indicators exceed the warning threshold for 10, 20, and 30 minutes respectively, the number of indicators is 3, and the durations of exceeding the threshold are 10, 20, and 30 minutes respectively. According to preset rules, the warning level can be divided into three levels: low, medium, and high. The corresponding warning level is determined based on the number of indicators and the duration of exceeding the threshold. For example, if the names of the target temperature characteristic indicators are temperature 1, temperature 2, and temperature 3, and the warning levels are low, medium, and high respectively, then based on this information, a fault warning message for the corresponding air cooler can be generated. For instance, if temperature 1 of air cooler 1 exceeds the warning threshold, the warning level is low.

[0114] The present invention also discloses a generator air cooler fault early warning system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the generator air cooler fault early warning method as described in any of the above embodiments.

[0115] The present invention also discloses a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the generator air cooler fault early warning method described in any of the above embodiments.

[0116] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of protection of this application is limited to these examples; under the concept of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of one or more embodiments of this application as described above, which are not provided in detail for the sake of brevity.

[0117] One or more embodiments in this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of this application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments in this application should be included within the protection scope of this application.

Claims

1. A method for early warning of generator air cooler faults, characterized in that, Includes the following steps: The unit monitoring system of the target generator set collects unit monitoring data of multiple air coolers in the target generator set. The unit monitoring data includes unit measurement point data, status timing data and external parameter data. The external parameter data is preprocessed into parameter assignment data; The data from the unit's measuring points were cleaned and supplemented. Based on preset temperature algorithm parameters and combined with the state time series data, all the unit measurement point data are calculated and processed to obtain multiple temperature characteristic values ​​of the air cooler; Temperature characteristic indices are generated by combining the formal parameter assignment data and the temperature characteristic values; The system combines the preset warning threshold and the temperature characteristic index to generate fault warning information corresponding to the air cooler. The preset temperature algorithm parameters include inlet and outlet temperature algorithm parameters and hot and cold air temperature algorithm parameters. The temperature feature values ​​include the calculated inlet temperature, calculated outlet temperature, inlet temperature range, outlet temperature range, hot and cold air temperature difference, and hot and cold air temperature difference dispersion of the air cooler. The process of calculating and processing all the unit's measuring point data based on preset temperature algorithm parameters and combined with the state time series data to obtain multiple temperature characteristic values ​​of the air cooler includes the following steps: According to the generator set number corresponding to the unit measurement point data, all the unit measurement point data are grouped into a first measurement point data group and a second measurement point data group. The unit measurement point data in the first measurement point data group and the second measurement point data group are traversed sequentially, and the unit measurement point data in each measurement point data group are sorted according to the measurement point data time axis; Based on the state time series data, data filtering is performed on the sorted first measurement point data group and the second measurement point data group respectively; The inlet and outlet temperature algorithm parameters are used to calculate and process the unit measuring point data in the first measuring point data group to obtain the inlet and outlet temperature calculation quantity and the inlet and outlet temperature range. The unit measuring point data in the second measuring point data group are calculated and processed using the cold and hot air temperature algorithm parameters to obtain the cold and hot air temperature difference and the dispersion of the cold and hot air temperature difference. The status time series data includes unit start-up duration and unit start-up delay. The unit measurement point data includes unit active power, unit start-up status, and unit start-up time. The data filtering based on the status time series data for the sorted first and second measurement point data groups includes the following steps: Filter out all unit measurement point data in the first and second measurement point data groups where the unit is in the off state; Determine whether the unit startup time is greater than or equal to the sum of the corresponding unit startup delay and the unit startup time; If the unit start-up time is less than the sum of the corresponding unit start-up delay and the unit start-up time, then the corresponding unit measurement point data will be filtered out. If the unit start-up time is greater than or equal to the sum of the corresponding unit start-up delay and the unit start-up time, then it is determined whether the corresponding active power of the unit is greater than a preset power threshold. If the active power of the corresponding unit is less than or equal to the power threshold, then the corresponding unit measurement point data is filtered out. If the active power of the corresponding unit is greater than the power threshold, then the corresponding unit measurement point data is retained; The unit's measurement data includes the air cooler inlet temperature and the air cooler outlet temperature. The step of calculating and processing the unit's measurement data in the first measurement data set using the inlet and outlet temperature algorithm parameters to obtain the calculated inlet and outlet temperatures and the inlet and outlet temperature range includes the following steps: According to the characteristic fields of the air cooler inlet temperature and the air cooler outlet temperature, all the unit measurement point data in the first measurement point data group are traversed, and all the air cooler inlet temperatures in the first measurement point data group are divided into inlet temperature group, and all the air cooler outlet temperatures in the first measurement point data group are divided into outlet temperature group. Rearrange all data in the inlet temperature group and the outlet temperature group in descending order; Based on the inlet and outlet temperature algorithm parameters, the average value of the first three air cooler inlet temperatures in the inlet temperature group is calculated as the inlet temperature calculation value, and the average value of the first three air cooler outlet temperatures in the outlet temperature group is calculated as the outlet temperature calculation value. Based on the inlet and outlet temperature algorithm parameters, the difference between the maximum and minimum values ​​of the air cooler inlet temperature in the inlet temperature group is calculated as the inlet temperature range, and the difference between the maximum and minimum values ​​of the air cooler outlet temperature in the outlet temperature group is calculated as the outlet temperature range.

2. The generator air cooler fault early warning method according to claim 1, characterized in that, The unit measurement data includes the air cooler cold air temperature and the air cooler hot air temperature. The calculation and processing of the unit measurement data in the second measurement data group using the cold and hot air temperature algorithm parameters to obtain the cold and hot air temperature difference and its dispersion includes the following steps: Based on each of the unit measurement point data, all the unit measurement point data in the second measurement point data group are traversed, and the air cooler cold air temperature and the air cooler hot air temperature belonging to the same measurement point in the second measurement point data group are divided into an air cooler temperature group. Based on the cold and hot air temperature algorithm parameters, the difference between the cold air temperature and the hot air temperature of the air cooler in each of the air cooler temperature groups is calculated to obtain the cold and hot air temperature difference in discrete data form. Calculate the data dispersion of all the hot and cold air temperature differences, and use the data dispersion as the dispersion of the hot and cold air temperature differences.

3. The generator air cooler fault early warning method according to claim 1, characterized in that, The data cleaning and data completion process for the unit's measuring point data includes the following steps: Iterate through all the unit's measurement point data and filter out abnormal data in the unit's measurement point data; Calculate the mean and standard deviation of all the unit measurement data; All unit measurement point data that differ from the average value by more than 3 times the standard deviation are removed; The data of the unit's measuring points after screening are supplemented in seconds according to the method of pre-complementation.

4. The generator air cooler fault early warning method according to claim 1, characterized in that, The process of generating fault warning information for the air cooler by combining the preset warning threshold and the temperature characteristic index includes the following steps: Compare all the temperature characteristic indicators with the preset warning thresholds one by one; If there is a target temperature feature index that exceeds the warning threshold, then extract the index name of the target temperature feature index; The number of the target temperature characteristic indicators and the duration of each target temperature characteristic indicator exceeding the warning threshold are counted. By combining the number of indicators and the market exceeding the standard, a warning level corresponding to the target temperature characteristic indicator is generated; By combining the indicator name and the warning level, a fault warning message corresponding to the air cooler is generated.

5. A generator air cooler fault early warning system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the generator air cooler fault early warning method as described in any one of claims 1 to 4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the generator air cooler fault early warning method according to any one of claims 1 to 4.

Images