A method for determining reasonable operating thresholds of key parameters of thermal power units

Abstract

The invention relates to a method for determining a reasonable operation threshold value of a key parameter of a thermal power unit. The method of the invention mines reasonable operation threshold values ​​of key operation parameters of the unit under different loads, back pressures, circulating water inlet temperatures, unit operation states and the like from historical data of unit operation by means of big data mining. The threshold range is close to normal values ​​of the unit operation parameters and includes almost all normal operation values. Compared with the original alarm values ​​and trip values ​​of a DCS, the threshold range has a stronger guiding effect on operation personnel. The method of the invention can establish a set of auxiliary operation monitoring system to assist operation personnel in monitoring. The system can monitor the operation state of key operation parameters of the unit in real time. Once the parameter operation index exceeds the reasonable threshold operation range, the system will send an alarm message to the operation personnel in real time, reminding the operation personnel in advance to adjust relevant operation parameters, eliminate operation defects in time, and ensure that the thermal power unit is always safely and stably operated.

Description

Technical Field

[0001] The present invention relates to the technical field of big data mining and determination of reasonable operating threshold values ​​of key parameters of thermal power units, and in particular to a method for determining reasonable operating threshold values ​​of key parameters of thermal power units based on big data mining. Background Art

[0002] The centralized control operation of unit units has become the main control method for large thermal power generating units today. When the thermal power generating units are in normal operation, the operators need to realize various monitoring and operations such as the start-up, shutdown, normal operation, monitoring and accident handling of the unit units in the centralized control room. The operation of modern thermal power generating units is a combination of complex boilers, steam turbines, generators and their auxiliary systems and complex automatic control systems. For operators, they must not only be familiar with the operating mechanisms of various systems such as boilers, steam turbines and generators, but also understand the control functions and control processes of the automatic control system. Therefore, the importance of operation management in various management of thermal power plants has become increasingly prominent.

[0003] Due to the special social nature of the thermal power unit industry, once the generator set is in operation, all equipment needs to operate stably 24 hours a day, and the operating personnel also need to be on duty 24 hours a day to conduct real-time monitoring and operation control. According to rough statistics, when the unit is operating normally, the operating personnel need to monitor more than 100 pictures and more than 2,000 operating parameters. Once a parameter is abnormal and the operating personnel fail to discover and adjust it in time, it will cause the system's multi-parameter correlation abnormality, which will cause equipment shutdown and damage. In severe cases, it will cause the unit to shut down, affecting the safety and economy of power plant operation, and even threaten the safety of local power grids, causing serious social impact.

[0004] At present, through investigations with operators of many power plants, it is found that operators have the following difficulties and pain points when monitoring and adjusting parameters:

[0005] (1) There is a large difference between the existing operating alarm values ​​and trip values ​​of key parameters and the normal operating values ​​of each parameter, and their guiding role for the normal operation of the unit is not strong.

[0006] (2) The operating personnel are getting younger. Many experienced operating personnel have left their operating positions for various reasons, while new employees who have been employed for one or two years are mostly responsible for monitoring tasks during actual unit operations. Due to their lack of experience, young employees are prone to not being clear about the monitoring parameter thresholds, which may lead to accidents.

[0007] (3) There are a large number of monitoring screens and parameters. It is easy to miss a screen or a monitoring parameter during monitoring. Often these parameters will deteriorate and cause a fault.

[0008] (4) Because operating personnel often work in shifts of five to three or five to four, especially the night shift and the night shift, they are prone to lack of energy or negligence when performing operation monitoring tasks, resulting in the continuous deterioration of important operating parameters of the unit without being discovered, causing failures.

[0009] (5) Due to the changes in parameters such as unit load and ambient temperature, the normal operating values ​​of some important operating parameters vary greatly, which is not easy to attract the attention of operating personnel, or the operation monitoring personnel do not understand the range of changes in their normal operating values ​​and blindly adjust them, causing faults.

[0010] Therefore, its improvement and innovation are imperative.

[0011] In view of the problems and difficulties that are prone to occur to operation monitoring personnel during the operation of the above-mentioned thermal power units, there is an urgent need for an auxiliary monitoring system for operators. Through big data mining technology, the normal operating threshold range of the key operating parameters of the unit can be determined, and the unit operating parameters can be automatically monitored to see if they are within the normal threshold range. If the operating parameters exceed the normal threshold range, the system will automatically alarm and prompt the operator to make relevant adjustments, eliminate defects in a timely manner, and ensure that the thermal power unit always maintains safe and stable operation. Summary of the invention

[0012] In view of the above situation, in order to overcome the defects of the prior art, the purpose of the present invention is to provide a method for determining the reasonable operating threshold values ​​of key parameters of a thermal power unit, which can effectively solve the problem of determining the normal operating threshold range of the key operating parameters of the unit, thereby reasonably monitoring the key parameters of the thermal power plant and ensuring that the thermal power unit always maintains safe and stable operation.

[0013] The technical solution provided by the present invention is:

[0014] A method for determining a reasonable operating threshold value of key parameters of a thermal power unit comprises the following steps:

[0015] Step 1: Determine the key operating parameters of the thermal power unit that need to be monitored;

[0016] Parameters include but are not limited to the following: main steam pressure and temperature, reheat steam pressure and temperature, high-pressure cylinder exhaust pressure and temperature, medium-pressure cylinder exhaust pressure and temperature, low-pressure cylinder exhaust pressure, pressure and temperature of each extraction section, turbine regulating stage pressure, inlet and outlet water temperature and drain temperature of each heater, bearing temperature, shaft vibration and bearing vibration of each bearing, bearing metal temperature, end difference of each condenser, characteristic flow area of ​​each section of the turbine, lubricating oil temperature and pressure, lubricating oil pump current, constant cooling water pump current, open water pump current, closed water pump current, vacuum pump current, condensate pump current, electric feed water pump current, boiler water circulation pump current, circulating water pump current, fire-resistant oil pump current, induced draft fan current, forced draft fan current, primary fan current, current of each coal mill, generator excitation power, plant power, generator excitation current and voltage, temperature difference between upper and lower cylinders of high-pressure cylinder, temperature difference between upper and lower cylinders of medium-pressure cylinder, and expansion difference of low-pressure cylinder;

[0017] Step 2: Export the two-year historical operation data of the above related parameters from the DCS system, with a data interval of 1 second, and store it externally;

[0018] Step 3: Preprocess the exported historical data;

[0019] The historical data is preprocessed by using the method of batch data cleaning. The preprocessing methods include:

[0020] (1) For missing data, the missing value is filled by taking the average of the values ​​of the three consecutive moments before the missing value;

[0021] (2) For mutation data, the mutation data is directly removed and the data is supplemented by the average value of the three consecutive values ​​before the moment;

[0022] Step 4: Determine the threshold screening boundary conditions of each monitoring parameter;

[0023] Each parameter does not need to be monitored all the time, but only needs to be monitored under certain boundary conditions. According to the parameters that need to be monitored, it can be divided into the following three monitoring boundary conditions: after the unit is connected to the grid, after various pumps are started, and after various fans are started;

[0024] The monitoring parameters after the unit is connected to the grid include: main steam pressure and temperature, reheat steam pressure and temperature, high pressure cylinder exhaust pressure and temperature, medium pressure cylinder exhaust pressure and temperature, low pressure cylinder exhaust pressure, pressure and temperature of each extraction section, inlet and outlet water temperature and drain temperature of each heater, bearing temperature, shaft vibration and bearing vibration of each bearing, end difference of each condenser, characteristic flow area of ​​each section of the turbine, generator excitation power, plant power, generator excitation current and voltage, coal mill current;

[0025] Monitoring parameters of various pumps after starting: lubricating oil temperature and pressure, cooling water pump current, open water pump current, closed water pump current, vacuum pump current, condensate pump current, electric feed water pump current, boiler water circulation pump current, circulating water pump current;

[0026] After starting various fans: induced draft fan current, forced draft fan current, primary fan current;

[0027] Step 5: Determine the threshold of each monitoring parameter to filter the associated parameters;

[0028] The boundary conditions of the threshold screening associated parameters of each monitoring parameter can be divided into the following categories:

[0029] (1) Whether it is related to the load: 1 indicates that it is related to the load, and 0 indicates that it is not related to the load;

[0030] (2) Whether back pressure correction is required: whether back pressure correction is not required, indicated by 0; back pressure correction is required, indicated by 1; circulating water temperature correction is required, indicated by 2;

[0031] (3) Whether it is grid-connected: It is mainly divided into non-grid-connected state, represented by 0, and grid-connected state, represented by 1;

[0032] The above three boundary conditions are combined with the actual unit operation considerations and divided into the following 8 situations:

[0033] (1) Independent of load, no back pressure correction is required, non-grid-connected state, represented by 000, related parameters include:

[0034] Lubricating oil pressure and temperature, lubricating oil pump current, constant cooling water pump current, condensate pump current, closed water pump current, circulating water pump current, induced draft fan current, forced draft fan current, primary fan current;

[0035] (2) It is independent of the load and does not require back pressure correction. The grid-connected state is indicated as 001. The relevant parameters include: the temperature difference between the upper and lower cylinders of the high-pressure cylinder, the temperature difference between the upper and lower cylinders of the medium-pressure cylinder, the expansion difference of the low-pressure cylinder, the vacuum pump current, the fire-resistant oil pump current, the main steam temperature, the reheat steam temperature, the cylinder expansion deviation on both sides of the high and medium pressure cylinders, the bearing return oil temperature, the bearing metal temperature, the lubricating oil pressure and temperature, the lubricating oil pump current, the constant cooling water pump current, the condensate pump current, the closed water pump current, the circulating water pump current, the induced draft fan current, the forced draft fan current, the primary fan current, and the temperature of each steam extraction section;

[0036] (3) Independent of load, back pressure correction is required, non-grid connected state, represented by 010, related parameters include: vacuum pump current;

[0037] (4) Independent of load, back pressure correction is required, grid-connected state, represented by 011, related parameters include: vacuum pump current;

[0038] (5) Related to the load, no back pressure correction is required, grid-connected state, represented by 101, related parameters include: characteristic flow area of ​​each section of the turbine, shaft vibration and bearing vibration of each bearing, main steam pressure, high-pressure cylinder exhaust pressure, reheat steam pressure, intermediate-pressure cylinder exhaust pressure, and pressure of each extraction section;

[0039] (6) Related to the load, back pressure correction is required, grid connection status, represented by 111, and related parameters include: condenser end difference;

[0040] (7) Related to the load, circulating water temperature correction is required, non-grid-connected state, represented by 120, related parameters include: turbine regulating stage pressure;

[0041] (8) Related to the load, requiring circulating water temperature correction, grid-connected status, represented by 121, and related parameters include: turbine regulating stage pressure, characteristic flow area of ​​each section of the turbine;

[0042] For load-related situations, the load interval is set every 10MW from 20% rated load to 100% rated load; for back pressure correction, a back pressure interval is set from 3kPa to 12kPa, and a back pressure interval is set every 0.2kPa; for circulating water temperature correction, a temperature interval is set from 3℃ to 35℃, and a temperature interval is set every 1℃, and a total of 34 circulating water temperature points can be set; for grid-connected status, it is divided into two states: grid-connected status and non-grid-connected status; each monitoring parameter establishes a conditional grid screened by the threshold value associated with the parameter, and the conditional grid is stored in the database;

[0043] Step 6: According to the boundary conditions of each monitoring parameter set in step 4 and step 5 and the associated parameters of the threshold screening, data mining is carried out from the original historical data after preprocessing, and the maximum value, minimum value, average value, and standard deviation of each threshold screening associated parameter of the parameter in the two-year history are obtained under the premise of meeting each boundary condition, and stored under the corresponding condition grid;

[0044] Step 7: Determine reasonable operating thresholds for monitoring parameters;

[0045] Determined as ((μ-3σ)×(1-β), (μ+3σ)×(1+β)); is the reasonable threshold range of the monitoring parameter, where μ is the mean value of the threshold screening associated parameter, σ is the standard deviation of the threshold screening associated parameter, and β is the broadening coefficient;

[0046] Calculate the reasonable threshold range of each monitoring parameter in each condition grid and store it in the condition grid of the corresponding monitoring parameter;

[0047] Step 8: Calculate the reasonable threshold range of monitoring parameters in real time;

[0048] According to the unit operation status, combined with the actual load, actual back pressure, actual circulating water inlet temperature, and grid connection status, the actual reasonable operation threshold range of the real-time monitoring parameters can be calculated through linear interpolation, and the key operation parameters of the thermal power unit can be monitored through the calculated actual reasonable operation threshold range. After the reasonable operation threshold of the monitoring parameter is calculated in real time, if the operating parameter exceeds the reasonable threshold range, the system will immediately send an alarm to the operating personnel, reminding the operating personnel to pay attention to the exceeding parameters in time and adjust the parameter value to a reasonable operation threshold range.

[0049] The method of the present invention mines reasonable operating thresholds of key operating parameters of the unit under different loads, back pressures, circulating water inlet temperatures, unit operating states and other conditions from historical data of unit operation by means of big data mining. The threshold range is close to normal values ​​of the unit operating parameters and includes almost all normal operating values. Compared with the original alarm values ​​and trip values ​​of the DCS, it has a stronger guiding effect on the operating personnel. The method of the present invention can establish a set of auxiliary operation monitoring system to assist the operating personnel in monitoring. The system can monitor the operating state of the key operating parameters of the unit in real time. Once the parameter operating index exceeds the reasonable threshold operating range, it will send an alarm message to the operating personnel in real time, reminding the operating personnel in advance to adjust the relevant operating parameters, eliminate the operating defects in time, and ensure the safe and stable operation of the thermal power unit at all times. DETAILED DESCRIPTION

[0050] The specific implementation modes of the present invention are further described in detail below in conjunction with the embodiments.

[0051] The present invention provides a method for determining a reasonable operating threshold value of key parameters of a thermal power unit, comprising the following steps:

[0052] Step 1: Determine the key operating parameters of the thermal power unit that need to be monitored;

[0053] The method can be applied to the key operating parameters of the steam turbine and its auxiliary systems, boiler and its auxiliary systems, and generator steam turbine auxiliary systems of thermal power units, including but not limited to the following: main steam pressure and temperature, reheat steam pressure and temperature, high-pressure cylinder exhaust pressure and temperature, medium-pressure cylinder exhaust pressure and temperature, low-pressure cylinder exhaust pressure, pressure and temperature of each extraction section, turbine regulating stage pressure, inlet and outlet water temperature and drain temperature of each heater, bearing temperature, shaft vibration and bearing vibration of each bearing, bearing metal temperature, and each condenser end difference, characteristic flow area of ​​each section of the steam turbine, lubricating oil temperature and pressure, lubricating oil pump current, constant cooling water pump current, open water pump current, closed water pump current, vacuum pump current, condensate pump current, electric feed water pump current, boiler water circulation pump current, circulating water pump current, fire-resistant oil pump current, induced draft fan current, forced draft fan current, primary fan current, current of each coal mill, generator excitation power, plant power, generator excitation current and voltage, temperature difference between upper and lower cylinders of high-pressure cylinder, temperature difference between upper and lower cylinders of medium-pressure cylinder, expansion difference of low-pressure cylinder;

[0054] Step 2: Export the two-year historical operation data of the above related parameters from the DCS system, with a data interval of 1 second, and store it externally;

[0055] Step 3: Preprocess the exported historical data;

[0056] The historical data is preprocessed by using the method of batch data cleaning. The preprocessing methods include:

[0057] (1) For missing data, the missing value is filled by taking the average of the values ​​of the three consecutive moments before the missing value;

[0058] (2) For mutation data, the mutation data is directly removed and the data is supplemented by the average value of the three consecutive values ​​before the moment;

[0059] Step 4: Determine the threshold screening boundary conditions of each monitoring parameter;

[0060] Each parameter does not need to be monitored all the time, but only needs to be monitored under certain boundary conditions. According to the parameters that need to be monitored, it can be divided into the following three monitoring boundary conditions: after the unit is connected to the grid, after various pumps are started, and after various fans are started;

[0061] The monitoring parameters after the unit is connected to the grid include: main steam pressure and temperature, reheat steam pressure and temperature, high pressure cylinder exhaust pressure and temperature, medium pressure cylinder exhaust pressure and temperature, low pressure cylinder exhaust pressure, pressure and temperature of each extraction section, inlet and outlet water temperature and drain temperature of each heater, bearing temperature, shaft vibration and bearing vibration of each bearing, end difference of each condenser, characteristic flow area of ​​each section of the turbine, generator excitation power, plant power, generator excitation current and voltage, coal mill current;

[0062] Monitoring parameters of various pumps after starting: lubricating oil temperature and pressure, cooling water pump current, open water pump current, closed water pump current, vacuum pump current, condensate pump current, electric feed water pump current, boiler water circulation pump current, circulating water pump current;

[0063] After starting various fans: induced draft fan current, forced draft fan current, primary fan current;

[0064] Step 5: Determine the threshold of each monitoring parameter to filter the associated parameters;

[0065] With the changes in the unit's operating load, back pressure, circulating water inlet temperature and unit operating status, the normal operating values ​​of different operating parameters will change significantly. However, different parameters may be associated with one or all of the unit load, circulating water inlet temperature and unit operating status, or none of them may be associated. This requires us to set different screening conditions according to the different states of each parameter. The boundary conditions of the threshold screening associated parameters of each monitoring parameter can be divided into the following types:

[0066] (1) Whether it is related to the load: 1 indicates that it is related to the load, and 0 indicates that it is not related to the load;

[0067] (2) Whether back pressure correction is required: whether back pressure correction is not required, indicated by 0; back pressure correction is required, indicated by 1; circulating water temperature correction is required, indicated by 2;

[0068] (3) Whether it is grid-connected: It is mainly divided into non-grid-connected state, represented by 0, and grid-connected state, represented by 1;

[0069] The above three boundary conditions are combined with the actual unit operation considerations and divided into the following 8 situations:

[0070] (1) Independent of load, no back pressure correction is required, non-grid-connected state, represented by 000, related parameters include:

[0071] Lubricating oil pressure and temperature, lubricating oil pump current, constant cooling water pump current, condensate pump current, closed water pump current, circulating water pump current, induced draft fan current, forced draft fan current, primary fan current;

[0072] (2) It is independent of the load and does not require back pressure correction. The grid-connected state is indicated as 001. The relevant parameters include: the temperature difference between the upper and lower cylinders of the high-pressure cylinder, the temperature difference between the upper and lower cylinders of the medium-pressure cylinder, the expansion difference of the low-pressure cylinder, the vacuum pump current, the fire-resistant oil pump current, the main steam temperature, the reheat steam temperature, the cylinder expansion deviation on both sides of the high and medium pressure cylinders, the bearing return oil temperature, the bearing metal temperature, the lubricating oil pressure and temperature, the lubricating oil pump current, the constant cooling water pump current, the condensate pump current, the closed water pump current, the circulating water pump current, the induced draft fan current, the forced draft fan current, the primary fan current, and the temperature of each steam extraction section;

[0073] (3) Independent of load, back pressure correction is required, non-grid connected state, represented by 010, related parameters include: vacuum pump current;

[0074] (4) Independent of load, back pressure correction is required, grid-connected state, represented by 011, related parameters include: vacuum pump current;

[0075] (5) Related to the load, no back pressure correction is required, grid-connected state, represented by 101, related parameters include: characteristic flow area of ​​each section of the turbine, shaft vibration and bearing vibration of each bearing, main steam pressure, high-pressure cylinder exhaust pressure, reheat steam pressure, intermediate-pressure cylinder exhaust pressure, and pressure of each extraction section;

[0076] (6) Related to the load, back pressure correction is required, grid connection status, represented by 111, and related parameters include: condenser end difference;

[0077] (7) Related to the load, circulating water temperature correction is required, non-grid-connected state, represented by 120, related parameters include: turbine regulating stage pressure;

[0078] (8) Related to the load, requiring circulating water temperature correction, grid-connected status, represented by 121, and related parameters include: turbine regulating stage pressure, characteristic flow area of ​​each section of the turbine;

[0079] As shown in the following table:

[0080]

[0081] For load-related situations, the load interval is set every 10MW from 20% rated load to 100% rated load; for back pressure correction, a back pressure interval is set from 3kPa to 12kPa, and a back pressure interval is set every 0.2kPa; for circulating water temperature correction, a temperature interval is set from 3℃ to 35℃, and a temperature interval is set every 1℃, and a total of 34 circulating water temperature points can be set; for grid-connected status, it is divided into two states: grid-connected status and non-grid-connected status; each monitoring parameter establishes a conditional grid screened by the threshold value associated with the parameter, and the conditional grid is stored in the database;

[0082] The conditional grid is as follows:

[0083] The conditional grid established by the 000 and 001 screening conditions is shown in the following table. In each grid of the table, the maximum value, minimum value, average value, and standard deviation of each parameter that meets the conditions are filled through 2 years of historical operation data mining.

[0084] condition Grid connection Not connected to the grid Parameter 1 Parameter 2 … Parameter n

[0085] The conditional grid established by the 010 and 011 screening conditions is shown in the following table. In each grid of the table, the maximum value, minimum value, average value, and standard deviation of each parameter that meets the conditions are filled through 2 years of historical operation data mining.

[0086] condition Grid connection Not connected to the grid 3.0kPa 3.2kPa 3.4kPa … 11.6kPa 11.8kPa 12.0kPa

[0087] The conditional grid established by 101 screening conditions is shown in the following table. In each grid of the table, the maximum value, minimum value, average value, and standard deviation of each parameter that meets the conditions are filled through 2 years of historical operation data mining.

[0088] condition Grid connection Not connected to the grid 20%THA 20%THA+10MW 20%THA+20MW … THA-20MW THA-10MW THA

[0089] The conditional grid established by the 111 screening conditions is shown in the following table. In each grid of the table, the maximum value, minimum value, average value, and standard deviation of each parameter that meets the conditions are filled through 2 years of historical operation data mining.

[0090] Grid connection 3.0kPa 3.2kPa 3.4kPa … 11.6kPa 11.8kPa 12.0kPa 20%THA 20%THA+10MW 20%THA+20MW … THA-20MW THA-10MW THA

[0091] The conditional grid established by the 120 and 121 screening conditions is shown in the following table. In each grid of the table, the maximum value, minimum value, average value, and standard deviation of each parameter that meets the conditions are filled in through 2 years of historical operation data mining.

[0092] Grid-connected / Not grid-connected 3℃ 4℃ 5℃ … 33℃ 34℃ 35℃ 20%THA 20%THA+10MW 20%THA+20MW … THA-20MW THA-10MW THA

[0093] Step 6: According to the boundary conditions of each monitoring parameter set in step 4 and step 5 and the associated parameters of the threshold screening, data mining is carried out from the original historical data after preprocessing, and the maximum value, minimum value, average value, and standard deviation of each threshold screening associated parameter of the parameter in the two-year history are obtained under the premise of meeting each boundary condition, and stored under the corresponding condition grid;

[0094] Step 7: Determine reasonable operating thresholds for monitoring parameters;

[0095] According to the 3σ rule, for a normally distributed data set, the probability that the value of a normal variable falls outside the interval (μ-3σ, μ+3σ) is less than 0.003, so the interval range of (μ-3σ, μ+3σ) is regarded as the value range of the normal variable; for this patent, because the parameters to be monitored by the thermal power unit do not all conform to the characteristics of the normal distribution, the 3σ rule is appropriately improved and determined to be ((μ-3σ)×(1-β), (μ+3σ)×(1+β)); is a reasonable threshold range for monitoring parameters, where μ is the average value of the threshold screening associated parameter, σ is the standard deviation of the threshold screening associated parameter, and β is the broadening coefficient;

[0096] The value of the broadening coefficient β in step 7 is set according to the specific unit, and the value of β is between 5% and 15%. This value is determined by mining historical data of different types of units. In the technical solution of the present invention, β of 300MW-level units is 15%, β of 600MW-level units is 10%, and β of 1000MW-level units is 5%. This can not only reasonably envelop the normal operating values ​​of all parameters within a reasonable threshold range, but also avoid frequent system alarms causing interference to operating personnel.

[0097] It is set according to the specific unit (β is between 5% and 15%, which is determined by mining historical data of different types of units. In this patent, 300MW units take 15%, 600MW units take 10%, and 1000MW units take 5%). This can not only reasonably envelop the normal operating values ​​of all parameters within a reasonable threshold range, but also avoid frequent system alarms causing interference to operators. According to the above method, the reasonable threshold range of each monitoring parameter in each conditional grid can be calculated and stored in the conditional grid of the corresponding monitoring parameter;

[0098] Step 8: Calculate the reasonable threshold range of monitoring parameters in real time;

[0099] According to the unit operation status, combined with the actual load, actual back pressure, actual circulating water inlet temperature, and grid connection status, the actual reasonable operation threshold range of the real-time monitoring parameters can be calculated through linear interpolation, and the key operation parameters of the thermal power unit can be monitored through the calculated actual reasonable operation threshold range. After the reasonable operation threshold of the monitoring parameter is calculated in real time, if the operating parameter exceeds the reasonable threshold range, the system will immediately send an alarm to the operator to remind the operator to pay attention to the exceeding parameters in time and adjust the parameter value to a reasonable operation threshold range.

[0100] The specific calculation method is explained by taking the condenser end difference as an example. The associated parameters of the condenser end difference are related to load, back pressure correction, and grid connection status, and its screening conditions can be simplified to 111.

[0101] ① Obtain the conditional grid of the condenser end difference from steps 5, 6 and 7, and select the actual reasonable operating thresholds in the conditional grid under different conditions from historical data.

[0102] ② Determine the condition parameters of the unit during actual operation. In the actual operation of the unit, the unit is in grid-connected state at this moment, the load is POWER, and the back pressure is P.

[0103] ③ The screening condition of the condenser end difference can be simplified to 111. According to the screening condition code 111, select the table numbered 111 in the database.

[0104] ④ In table 111, find the upper and lower limits closest to the load POWER, which are POWER_down and POWER_up respectively; find the upper and lower limits closest to the back pressure P in table 111, which are P_down and P_up respectively. The filtered condition grid is shown in the following table. The search method is the mutual comparison method, that is, the actual load value or vacuum value is compared with the load value and vacuum value in the table in turn, so as to find the upper and lower limits closest to the actual value.

[0105]

[0106] ⑤ According to the above table, under the P_down condition, the load is linearly interpolated. The lower and upper thresholds under different loads are linearly interpolated using the following formula to obtain the reasonable operating threshold ranges E1 and E2 under the P_down condition.

[0107]

[0108] ⑥ Similarly, under the P_up condition, the load is linearly interpolated, and the lower and upper thresholds under different loads are linearly interpolated using the following formula to obtain the reasonable operating threshold ranges F1 and F2 under the P_up condition.

[0109]

[0110] ⑦ For the reasonable operation thresholds obtained under P_down and P_up conditions, perform linear interpolation on the back pressure. The lower and upper thresholds under different back pressures are linearly interpolated using the following formula to obtain the upper and lower limits G1 and G2 of the reasonable operation thresholds that meet both load and back pressure conditions.

[0111]

[0112] ⑧ The upper limit G1 and lower limit G2 of the reasonable operation threshold are the reasonable operation threshold range of the condenser end difference under the actual operation state, load and operation back pressure. The calculation and determination steps of the reasonable operation threshold range of other operation parameters are shown in ① to ⑦, which will not be described in detail here.

[0113] After the system is deployed, it can be started regularly, automatically mining the thresholds in historical data and updating the reasonable operating threshold range in the condition grid in real time. According to the changes in the unit operation, the reasonable operating threshold range is adjusted in real time to keep up with the actual operation of the unit, and the guiding role is strong. After the system is deployed, each time it is started, steps 3 to 7 can be automatically repeated to complete the update of the reasonable operating threshold of the parameters and guide the adjustment of the operating parameters in real time.

[0114] The present invention has achieved the same or similar effects through practical application, and the application examples are as follows:

[0115] Application Example 1:

[0116] Take a 1000MW unit of a Shanghai steam turbine plant as an example. The unit was put into operation in 2016. The vibration parameter monitoring of the unit has always been the focus of the plant operators. The bearing vibration has always been 50μm as the qualified value, 80μm as the adjustment value, and the shaft vibration has always been 120μm as the alarm value, and 250μm as the trip value. The operators have a heavy monitoring task and often need to pull historical curves to analyze the changing trend of bearing vibration.

[0117] Since the system was put into use in March 2020, through online big data mining and the reasonable threshold range of unit bearing vibration determined by this patent, the monitoring pressure of operators has been greatly reduced, and it is highly applicable and instructive. The following takes the X-direction vibration of bearing No. 3 as an example for introduction.

[0118] The boundary condition for the X-direction vibration of bearing No. 3 is determined to be the unit grid connection, and the only parameter associated with it is the unit load. Therefore, according to the requirements of this method, the load is divided into intervals and historical data is mined online. The reasonable threshold range of this parameter can be obtained as shown in the following table.

[0119]

[0120] After the system was put into operation, the reasonable threshold range of the vibration value of bearing No. 3 in the X direction (other parameters are similar) was determined. This threshold range is closely around the normal operation value of the parameter and is far away from the alarm value of the unit. It can be automatically monitored by the program within the reasonable threshold range, and the parameter degradation trend can be discovered in advance and handled in time.

[0121] On July 3, 2020, the system showed that the vibration of bearing No. 3 in the X direction exceeded the reasonable operating threshold range and triggered an alarm to notify the operator. The operator observed the vibration value of bearing No. 3 in the X direction in time through the alarm and found that the vibration of bearing No. 3 in the X direction continued to rise. After real-time analysis of the cause, the vibration of bearing No. 3 in the X direction was controlled and stopped rising by adjusting the lubricating oil temperature. This alarm reminded the operator in time and avoided the continuous increase of the vibration parameters of the unit. The operation adjustment measures were taken to eliminate the trend of parameter deterioration, so that the unit avoided an unplanned shutdown caused by large bearing vibration, which greatly guaranteed the safety and economy of the unit.

[0122] Application Example 2:

[0123] Take a 600MW unit in a Harbin steam turbine plant as an example. The unit underwent a flow-through modification in 2019. The monitoring of the thermal system parameters of the unit has always been the focus of monitoring by the plant's operators. Generally, the threshold value of the thermal system parameter monitoring is set as the design value of the parameter in the plant.

[0124] Since the system was put into use in May 2021, through online big data mining and the reasonable threshold range of the unit thermal system parameters determined by this patent, the monitoring pressure of the operating personnel has been greatly reduced, and the applicability and guidance are extremely strong. The following is an introduction using the unit condenser end difference as an example.

[0125] The power plant generally uses a fixed design value as the threshold for monitoring the condenser end difference, and the design values ​​for different units are different. The design value of the condenser end difference of this unit is 4.3℃, and the power plant generally uses 4.3℃ as the threshold. Generally, it is normal if it does not exceed 4.3℃. However, due to changes in ambient temperature and the influence of unit load, the operating range of the condenser end difference of this unit under different operating conditions varies between 1.5℃ and 10℃. The operating personnel cannot figure out the actual reasonable range of the condenser end difference under different loads and different circulating water inlet temperatures, so they reduce the monitoring of this parameter, increasing the risk of condenser operation deterioration without being discovered.

[0126] The boundary condition of the condenser terminal difference is determined as the unit grid connection, and the parameters associated with it are the unit load and the circulating water inlet temperature. Therefore, according to the requirements of this method, the load and circulating water inlet temperature are divided into intervals and historical data are mined online to obtain the reasonable threshold range of this parameter as shown in the following table.

[0127] After the system was put into operation, the reasonable threshold range of the condenser end difference (other parameters are similar) was determined. The threshold range is closely around the normal operating value of the parameter, and the threshold range is differentiated under different loads and circulating water inlet temperatures, which is more reasonable and in line with reality. At the same time, it can be automatically monitored by the program within the reasonable threshold range. Under different loads, the threshold can be interpolated and calculated according to the actual load based on the threshold range inside the conditional grid. The final use effect is good, which greatly reduces the burden of monitoring by operators and improves the safety of unit operation.

[0128]

[0129]

[0130]

[0131] On March 22, 2022, the system showed that the condenser end difference of the unit exceeded the reasonable operating threshold range and triggered an alarm to notify the operating personnel. The operating personnel observed the continuous increase in the condenser end difference value through the alarm. Under the actual operating load and circulating water inlet temperature, the reasonable operating range of the condenser end difference is 2.0℃ to 4.4℃, and the actual condenser operating end difference continued to rise to 8.6℃ after exceeding the actual reasonable operating threshold. After the operating personnel discovered this phenomenon, they checked the circulating water system and vacuum system of the unit in real time, and carried out on-site inspections on site. It was found that the filter between the cooling tower and the circulating water pump was blocked by debris and plastic cloth, which greatly reduced the circulating water flow rate. As a result, the amount of circulating water entering the condenser was insufficient, causing the condenser end difference to increase. After the operating personnel promptly cleaned the blockage of the filter and cleaned the flow channel, the end difference of the unit condenser slowly returned to the normal operating range. This alarm promptly reminded the operating personnel and accurately gave the range of the condenser end difference exceeding the standard, and gave the deviation from the normal operating threshold. This enabled the operating personnel to carry out continuous monitoring of parameters that are often neglected, and quickly conduct problem troubleshooting, find out the ultimate cause of the problem and solve it, which is of great help to the economy of the unit.

Claims

1. A method for determining reasonable operating thresholds of key parameters of thermal power units, characterized in that: The following steps are involved: Step 1: Determine the key operating parameters of the thermal power unit that need to be monitored; Parameters include but are not limited to the following: main steam pressure and temperature, reheat steam pressure and temperature, high-pressure cylinder exhaust pressure and temperature, medium-pressure cylinder exhaust pressure and temperature, low-pressure cylinder exhaust pressure, pressure and temperature of each extraction section, turbine regulating stage pressure, inlet and outlet water temperature and drain temperature of each heater, bearing temperature, shaft vibration and bearing vibration of each bearing, bearing metal temperature, end difference of each condenser, characteristic flow area of ​​each section of the turbine, lubricating oil temperature and pressure, lubricating oil pump current, constant cooling water pump current, open water pump current, closed water pump current, vacuum pump current, condensate pump current, electric feed water pump current, boiler water circulation pump current, circulating water pump current, fire-resistant oil pump current, induced draft fan current, forced draft fan current, primary fan current, current of each coal mill, generator excitation power, plant power, generator excitation current and voltage, temperature difference between upper and lower cylinders of high-pressure cylinder, temperature difference between upper and lower cylinders of medium-pressure cylinder, and expansion difference of low-pressure cylinder; Step 2: Export the two-year historical operation data of the above related parameters from the DCS system, with one data point at an interval of 1 second, and store it externally; Step 3: Preprocess the exported historical data; The historical data is preprocessed by using the method of batch data cleaning. The preprocessing methods include: (1) For missing data, the missing value is filled by taking the average of the values ​​of the three consecutive moments before the missing value; (2) For mutation data, the mutation data is directly removed and the data is supplemented by the average value of the three consecutive values ​​before the moment; The mutation data refers to the data at this moment whose absolute value is obviously different from the absolute value of the data at the previous moment. Here, it is defined as "large difference" if the absolute value of the data at this moment is more than 20 times the absolute value of the data at the previous moment, and the data at this moment is defined as mutation data; Step 4: Determine the threshold screening boundary conditions of each monitoring parameter; Each parameter does not need to be monitored all the time, but only needs to be monitored under certain boundary conditions. According to the parameters that need to be monitored, it is divided into the following three monitoring boundary conditions: after the unit is connected to the grid, after various pumps are started, and after various fans are started; The monitoring parameters after the unit is connected to the grid include: main steam pressure and temperature, reheat steam pressure and temperature, high pressure cylinder exhaust pressure and temperature, medium pressure cylinder exhaust pressure and temperature, low pressure cylinder exhaust pressure, pressure and temperature of each extraction section, inlet and outlet water temperature and drain temperature of each heater, bearing temperature, shaft vibration and bearing vibration of each bearing, end difference of each condenser, characteristic flow area of ​​each section of the turbine, generator excitation power, plant power, generator excitation current and voltage, coal mill current; Monitoring parameters of various pumps after starting: lubricating oil temperature and pressure, cooling water pump current, open water pump current, closed water pump current, vacuum pump current, condensate pump current, electric feed water pump current, boiler water circulation pump current, circulating water pump current; After starting various fans: induced draft fan current, forced draft fan current, primary fan current; Step 5: Determine the threshold of each monitoring parameter to filter the associated parameters; The boundary conditions of the threshold screening associated parameters of each monitoring parameter are divided into the following categories: (1) Whether it is related to the load: 1 indicates that it is related to the load, and 0 indicates that it is not related to the load; (2) Whether back pressure correction is required: whether back pressure correction is not required, indicated by 0; back pressure correction is required, indicated by 1; circulating water temperature correction is required, indicated by 2; (3) Whether it is grid-connected: it is divided into non-grid-connected state, represented by 0, and grid-connected state, represented by 1; The above three boundary conditions are combined with the actual unit operation considerations and divided into the following 8 situations: (1) Independent of load, no back pressure correction is required, non-grid-connected state, represented by 000, related parameters include: Lubricating oil pressure and temperature, lubricating oil pump current, constant cooling water pump current, condensate pump current, closed water pump current, circulating water pump current, induced draft fan current, forced draft fan current, primary fan current; (2) It is independent of the load and does not require back pressure correction. The grid-connected state is indicated as 001. The relevant parameters include: the temperature difference between the upper and lower cylinders of the high-pressure cylinder, the temperature difference between the upper and lower cylinders of the medium-pressure cylinder, the expansion difference of the low-pressure cylinder, the vacuum pump current, the fire-resistant oil pump current, the main steam temperature, the reheat steam temperature, the cylinder expansion deviation on both sides of the high and medium pressure cylinders, the bearing return oil temperature, the bearing metal temperature, the lubricating oil pressure and temperature, the lubricating oil pump current, the constant cooling water pump current, the condensate pump current, the closed water pump current, the circulating water pump current, the induced draft fan current, the forced draft fan current, the primary fan current, and the temperature of each steam extraction section; (3) Independent of load, back pressure correction is required, non-grid connected state, represented by 010, related parameters include: vacuum pump current; (4) Independent of load, back pressure correction is required, grid-connected state, represented by 011, related parameters include: vacuum pump current; (5) Related to the load, no back pressure correction is required, grid-connected state, represented by 101, related parameters include: characteristic flow area of ​​each section of the turbine, shaft vibration and bearing vibration of each bearing, main steam pressure, high-pressure cylinder exhaust pressure, reheat steam pressure, intermediate-pressure cylinder exhaust pressure, and pressure of each extraction section; (6) Related to the load, back pressure correction is required, grid connection status, represented by 111, and related parameters include: condenser end difference; (7) Related to the load, circulating water temperature correction is required, non-grid-connected state, represented by 120, related parameters include: turbine regulating stage pressure; (8) Related to the load, requiring circulating water temperature correction, grid-connected status, represented by 121, and related parameters include: turbine regulating stage pressure, characteristic flow area of ​​each section of the turbine; For load-related situations, load intervals are set every 10MW from 20% rated load to 100% rated load; for back pressure correction, a back pressure interval is set from 3kPa to 12kPa, and a back pressure interval is set every 0.2kPa; for circulating water temperature correction, a temperature interval is set from 3℃ to 35℃, and a temperature interval is set every 1℃, with a total of 34 circulating water temperature points; for grid-connected status, it is divided into grid-connected status and non-grid-connected status; each monitoring parameter establishes a conditional grid screened by a threshold value associated with the parameter, and the conditional grid is stored in the database; Step 6: According to the boundary conditions of each monitoring parameter set in step 4 and step 5 and the associated parameters of the threshold screening, data mining is carried out from the original historical data after preprocessing, and the maximum value, minimum value, average value, and standard deviation of each threshold screening associated parameter of the parameter in the two-year history are obtained under the premise of meeting each boundary condition, and stored under the corresponding condition grid; Step 7: Determine reasonable operating thresholds for monitoring parameters; Determined as ((μ-3σ)×(1-β), (μ+3σ)×(1+β)); is the reasonable threshold range of the monitoring parameter, where μ is the mean value of the threshold screening associated parameter, σ is the standard deviation of the threshold screening associated parameter, and β is the broadening coefficient; Calculate the reasonable threshold range of each monitoring parameter in each conditional grid and store it in the conditional grid of the corresponding monitoring parameter; Step 8: Calculate the reasonable threshold range of monitoring parameters in real time; According to the operating conditions of the unit, combined with the actual load, actual back pressure, actual circulating water inlet temperature, and grid connection conditions, the actual reasonable operating threshold range of the real-time monitoring parameters can be calculated through linear interpolation, and the key operating parameters of the thermal power unit can be monitored through the calculated actual reasonable operating threshold range.

2. The method for determining the reasonable operation threshold of key parameters of a thermal power unit according to claim 1 is characterized in that: The value of the widening coefficient β in step 7 is set according to the specific unit. For a 300MW unit, β is 15%, for a 600MW unit, β is 10%, and for a 1000MW unit, β is 5%.

Images