A method and system for predicting power generation gas consumption of a gas turbine unit

Abstract

This invention relates to the field of energy engineering technology, specifically disclosing a method and system for predicting the power consumption of a gas turbine unit. The method includes: identifying a window of extremely slowly varying operating conditions through a monitoring system, constructing a scatter plot, identifying closed narrow hysteresis loops to determine nonlinear hysteresis effects, completing wall thermal inertia analysis and quantifying the total wall heat flow rate; correcting combustion temperature drift based on the total wall heat flow rate to obtain the true combustion temperature; constructing an air-fuel ratio correction model in the sensitive range of the slightly open gas regulating valve to achieve dynamic matching of gas flow; eliminating gas consumption distortion components to obtain hysteresis-free true gas consumption values, and finally obtaining hourly and daily gas consumption prediction ranges without dynamic interference through cumulative calculation and system deviation calibration. The system includes: a hysteresis identification module, a temperature correction module, an opening degree adjustment module, a gas consumption analysis module, and a prediction output module. The solution effectively improves the accuracy of gas consumption prediction under extremely slowly varying peak-shaving conditions.

Description

Technical Field

[0001] This invention relates to the field of energy engineering technology, and specifically to a method and system for predicting the power consumption of a gas turbine unit. Background Technology

[0002] In scenarios involving deep peak shaving in the power grid, gas turbine units frequently perform AGC (Automatic Gain Control) extremely slow regulation with a load change rate of only 1% to 3% of the rated load per minute. This type of operating condition lies between dynamic regulation and steady state; it is neither a step load change nor a completely static steady state point.

[0003] Existing generator power consumption prediction models almost universally treat this extremely slowly changing operating condition as an approximate steady state, directly applying the steady-state gas consumption curve for calculation. They completely fail to consider the nonlinear hysteresis effect caused by wall heat storage in the combustion chamber, flue gas passage, and heat exchange components during the slow increase or decrease of load.

[0004] This hysteresis causes the combustion temperature field to drift slowly rather than quickly follow the load, resulting in a dynamic mismatch in the air-fuel ratio within the range of slight opening of the gas regulating valve. Gas consumption will exhibit a narrow hysteresis loop with lower values ​​during load increases and higher values ​​during load decreases. The error is small on a single occasion, but after long-term continuous accumulation, it will cause a systematic deviation of 0.3% to 0.6% in hourly and daily gas consumption predictions, which cannot be corrected by conventional models.

[0005] Therefore, the present invention provides a method and system for predicting the power consumption of a gas turbine unit. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for predicting the power consumption of a gas turbine unit, so as to solve the above-mentioned background problems.

[0007] The objective of this invention can be achieved through the following technical solution: a method and system for predicting the power consumption of a gas turbine unit, comprising: Identify the extremely slow-changing operating condition window of the gas turbine unit and construct a scatter plot; determine whether a closed narrow hysteresis loop occurs. If it does, determine that a nonlinear hysteresis effect has occurred, and perform wall thermal inertia analysis to obtain the total wall heat flux. The theoretical steady-state combustion temperature is calculated using the steady-state thermodynamic balance method based on the total wall heat flux, and the combustion temperature drift is obtained by drift analysis. The actual combustion temperature is then obtained by superimposing the load conditions and the theoretical steady-state combustion temperature. The real-time opening degree of the gas regulating valve is obtained to determine whether it is in the slightly open sensitive range. If it is, an air-fuel ratio correction model is constructed based on the actual combustion temperature to obtain the actual air-fuel ratio, and a mismatch analysis is performed to obtain the theoretical gas flow rate, and the real-time opening degree is adjusted accordingly. Instantaneous power consumption is calculated based on theoretical gas flow rate; theoretical power consumption from historical steady-state database of gas turbine unit is retrieved to set hysteresis reference value, and distortion analysis is performed in conjunction with instantaneous power consumption to obtain actual gas consumption value; The gas consumption at a single point is calculated based on the actual gas consumption value, and then accumulated to obtain the total gas consumption at the hourly and daily levels. Finally, range calibration analysis is performed to obtain the predicted range of gas consumption at the hourly and daily levels.

[0008] Furthermore, the process of constructing a scatter plot involves: The actual power generation efficiency and instantaneous volumetric flow rate of the gas turbine unit at different acquisition time points within the extremely slow-changing operating condition window are integrated to obtain the actual power generation efficiency sequence and the instantaneous volumetric flow rate sequence. Obtain the theoretical steady-state flow rate sequence, and calculate the difference between the instantaneous volumetric flow rate sequence and the theoretical steady-state flow rate sequence to obtain the real-time gas consumption difference; A scatter plot was constructed with the actual power generation efficiency at the time of data collection as the horizontal axis and the real-time gas consumption difference as the vertical axis.

[0009] Furthermore, the process of determining the occurrence of nonlinear hysteresis effect is as follows: In the scatter plot, if the real-time gas consumption difference corresponding to the actual power generation efficiency during the load increase process is below the horizontal axis, and the real-time gas consumption difference corresponding to the actual power generation efficiency during the load decrease process is above the horizontal axis, then a closed narrow hysteresis loop is determined to have occurred. If a closed narrow hysteresis loop appears, it is determined that the gas turbine unit is experiencing a nonlinear hysteresis effect within the extremely slowly changing operating condition window.

[0010] Furthermore, the process of obtaining the combustion temperature drift by performing drift analysis is as follows: The total flue gas volume, comprehensive heat transfer coefficient, and flue gas specific heat capacity at constant pressure are obtained in real time and multiplied to calculate the total wall heat flow rate. The ratio of the total wall heat flow rate to the product calculation result is then used to obtain the combustion temperature drift.

[0011] Furthermore, the process of obtaining the overall heat transfer coefficient is as follows: Based on the gas turbine design manual, the heat transfer coefficient U1 and effective heat exchange area A1 of the combustion chamber, the heat transfer coefficient U2 and effective heat exchange area A2 of the flue gas passage, and the heat transfer coefficient U3 and effective heat exchange area A3 of the heat exchange component were obtained respectively. The comprehensive heat transfer coefficient is calculated by combining the area-weighted average method in engineering thermophysics.

[0012] Furthermore, the process of performing mismatch analysis is as follows: The air-fuel ratio mismatch is calculated by obtaining the difference between the actual air-fuel ratio and the design air-fuel ratio. Obtain the theoretical steady-state flow rate, calculate the ratio of the air-fuel ratio mismatch to the design air-fuel ratio, and multiply the ratio calculation result with the theoretical steady-state flow rate to calculate the gas compensation amount; The corrected theoretical gas flow rate is obtained by summing the gas compensation amount and the instantaneous volumetric flow rate. The theoretical gas flow rate generation command is input into the gas turbine control system, which then adjusts the real-time opening of the gas regulating valve according to the target flow rate in a closed-loop manner.

[0013] Furthermore, the actual air-fuel ratio is obtained as follows: The current methane volume is obtained based on the real-time component parameters of the gas collected by the monitoring system. The real-time theoretical air volume is obtained by analyzing and calculating the current methane volume and the chemical reaction formula for the complete combustion of methane. The real-time total air volume is obtained by using a monitoring system, and the actual air-fuel ratio is calculated by comparing the real-time total air volume with the real-time theoretical air volume.

[0014] Furthermore, the process of setting the hysteresis reference value for load increase and decrease conditions is as follows: Obtain theoretical power consumption, unit base value, and narrow-range offset threshold; The difference between the unit base value and the narrow-range offset threshold is calculated, and the result of the difference calculation is multiplied by the theoretical power consumption of the generator to obtain the hysteresis base value for load increase. The unit base value and the narrow-range offset threshold are accumulated and calculated. The result of the accumulated calculation is multiplied with the theoretical power consumption to obtain the hysteresis base value for load reduction.

[0015] Furthermore, the process of performing range calibration analysis is as follows: The time-based and daily-based reference gas consumption values ​​predicted by the steady-state gas consumption prediction model under extremely slowly changing operating conditions are retrieved and the deviation range is calculated to obtain the time-based and daily-based deviation ranges. The final gas consumption prediction range for the hour / day is calculated by multiplying the actual gas consumption value by the maximum and minimum values ​​of the hourly / dayly reference deviation range.

[0016] A predictive system for the power consumption of a gas turbine unit includes the following: Hysteresis identification module: Identifies the extremely slow-changing operating condition window of the gas turbine unit and constructs a scatter plot; determines whether a closed narrow hysteresis loop occurs. If it does, it determines that a nonlinear hysteresis effect has occurred and performs wall thermal inertia analysis to obtain the total wall heat flux. Temperature correction module: The theoretical steady-state combustion temperature is calculated based on the total wall heat flux using the steady-state thermodynamic balance method. Drift analysis is performed to obtain the combustion temperature drift. The actual combustion temperature is obtained by superimposing the load conditions and the theoretical steady-state combustion temperature. Opening adjustment module: acquires the real-time opening of the gas regulating valve, determines whether it is in the slightly open sensitive range; if so, it constructs an air-fuel ratio correction model based on the actual combustion temperature to obtain the actual air-fuel ratio, performs mismatch analysis to obtain the theoretical gas flow rate, and adjusts the real-time opening. Gas consumption analysis module: Calculates instantaneous power consumption based on theoretical gas flow rate; retrieves theoretical power consumption from historical steady-state database of gas turbine unit to set hysteresis benchmark value, and performs distortion analysis in conjunction with instantaneous power consumption to obtain the actual gas consumption value; Prediction output module: Calculates the gas consumption at a single point based on the actual gas consumption value, accumulates it to obtain the total gas consumption at the hourly and daily levels; and performs range calibration analysis to obtain the final gas consumption prediction range for the hour and day.

[0017] The beneficial effects of this invention are as follows: 1. Identify the extremely slow-changing operating condition window of the gas turbine unit and construct a scatter plot; determine whether a closed narrow hysteresis loop occurs. If so, determine that a nonlinear hysteresis effect has occurred, and perform wall thermal inertia analysis to obtain the total wall heat flow rate; identify the nonlinear hysteresis effect, quantify the heat exchange caused by wall thermal inertia, and provide a reliable quantitative basis for subsequent hysteresis correction; calculate the theoretical steady-state combustion temperature based on the total wall heat flow rate using the steady-state thermodynamic balance method, and perform drift analysis to obtain the combustion temperature drift; obtain the actual combustion temperature by superimposing the load condition and the theoretical steady-state combustion temperature; eliminate the combustion temperature drift caused by wall thermal inertia, and make the actual combustion temperature change synchronously with the AGC extremely slow-changing load command to establish a hysteresis-free temperature benchmark; obtain the real-time opening of the gas regulating valve and determine whether it is in the slightly open sensitive range; if so, construct an air-fuel ratio correction model based on the actual combustion temperature to obtain the actual air-fuel ratio, perform mismatch analysis to obtain the theoretical gas flow rate, and adjust the real-time opening; avoid flow mismatch through dynamic air-fuel ratio correction to ensure that the unit is always in the optimal combustion state.

[0018] 2. Instantaneous power consumption is calculated based on theoretical gas flow rate; theoretical power consumption from the historical steady-state database of the gas turbine unit is used to set a hysteresis baseline value, and distortion analysis is performed in conjunction with instantaneous power consumption to obtain the true gas consumption value; the narrow hysteresis distortion component caused by wall thermal inertia is effectively removed to obtain pure true gas consumption without hysteresis interference, which truly reflects the gas consumption level of the unit itself; the gas consumption at a single point is calculated based on the true gas consumption value, and the total gas consumption at the hourly and daily levels is accumulated; range calibration analysis is performed to obtain the final gas consumption prediction range for the hourly and daily levels; the total gas consumption at the hourly and daily levels is accumulated, and a gas consumption prediction range without dynamic hysteresis and error accumulation is given, which greatly improves the accuracy and practicality of gas consumption prediction under extremely slow-changing operating conditions. Attached Figure Description

[0019] The invention will now be further described with reference to the accompanying drawings.

[0020] Figure 1 This is a flowchart of the steps in a method for predicting the power consumption of a gas turbine unit according to the present invention; Figure 2 This is a logic diagram for determining whether the gas regulating valve is in the slightly open sensitive range in this invention. Figure 3 This is a system module diagram of a predictive system for the power consumption of a gas turbine unit according to the present invention. Detailed Implementation

[0021] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0022] Example 1 like Figures 1-2 As shown, a method for predicting the power consumption of a gas turbine unit includes: Step 1: Use a monitoring system to monitor the gas turbine unit to identify the extremely slow-changing operating condition window, and collect real-time gas consumption difference to construct a scatter plot; determine whether a closed narrow hysteresis loop appears based on the scatter plot. If it appears, it is determined that a nonlinear hysteresis effect has occurred, and then perform wall thermal inertia analysis to obtain the total wall heat flux. In step one, the process of using a monitoring system to monitor the gas turbine unit to identify the extremely slowly varying operating condition window and collecting real-time gas consumption difference sequences to construct a scatter plot is as follows: Select a stable gas turbine unit to deploy the monitoring system; It should be noted that a stable gas turbine unit must meet the following conditions: the gas turbine unit is in healthy operation with no alarms or defects affecting performance; key metering instruments such as gas flow, thermocouples, and power transmitters are all within their valid calibration period to ensure reliable data sources; and the gas turbine unit control system has a high-speed data acquisition and storage interface. A monitoring system is used to obtain AGC (Automatic Generation Control) commands and the actual load change rate of the gas turbine unit in real time; When the monitoring system receives an AGC command from the power grid within the range of 1%Pe / min to 3%Pe / min, it determines that the unit has entered the target regulation mode and immediately activates the extremely slow variable operating condition buffer of the monitoring system. Where Pe is the rated load of the gas turbine unit; After the extremely slow variable operating condition buffer is activated, the monitoring system continuously compares the real-time load change rate of the gas turbine unit. If the actual load change rate is within the range of 1%~3%Pe / min, then this period is marked as the extremely slow change condition window; Within the extremely slow-changing operating condition window, the actual power generation efficiency, instantaneous volumetric flow rate, and representative temperature measurement points on key components such as the combustion chamber, flue gas passage, and heat exchange components of the gas turbine unit are collected. It should be noted that instantaneous volumetric flow rate refers to the actual volumetric flow rate of gas consumed by the gas turbine at a certain instant. The actual power generation efficiency and instantaneous volumetric flow rate at different acquisition time points within the extremely slow-changing operating condition window are integrated to obtain the actual power generation efficiency sequence and the instantaneous volumetric flow rate sequence. Obtain the theoretical steady-state flow rate corresponding to the actual power generation efficiency from the historical steady-state database of the gas turbine unit, and integrate them to obtain the theoretical steady-state flow rate sequence; The real-time gas consumption difference is obtained by calculating the difference between the instantaneous volumetric flow rate sequence and the theoretical steady-state flow rate sequence. It should be noted that the real-time gas consumption difference sequence reflects the deviation between the actual gas consumption and the steady-state theoretical value; its regular change of negative for load increase and positive for load decrease intuitively reflects the energy absorption or release lag effect caused by the thermal inertia of the metal wall. The actual power generation efficiency and real-time gas consumption difference at different collection time points within the extremely slow-changing operating condition window are obtained. A scatter plot is constructed with the actual power generation efficiency as the horizontal axis and the real-time gas consumption difference as the vertical axis. In step one, the process of determining whether a closed narrow hysteresis loop appears based on the scatter plot is as follows: If it does, it is determined that a nonlinear hysteresis effect has occurred, and the process of performing wall thermal inertia analysis to obtain the total wall heat flux is as follows: In the scatter plot, if the real-time gas consumption difference corresponding to the actual power generation efficiency during the load increase process is below the horizontal axis, and the real-time gas consumption difference corresponding to the actual power generation efficiency during the load decrease process is above the horizontal axis, then a closed narrow hysteresis loop is determined to have occurred. If a closed narrow hysteresis loop appears, it is determined that the gas turbine unit is experiencing a nonlinear hysteresis effect within the extremely slowly changing operating condition window. The total heat capacity of the combustion chamber is calculated by multiplying the specific heat capacity, density, and effective heat exchange volume of the combustion chamber material. Temperature data of combustion room temperature measuring points at different acquisition time points are obtained. Starting from the extremely slowly changing operating condition window, and with the calculation step size set to N, the termination point is obtained. The temperature difference between the termination point temperature data and the starting point temperature data is calculated. The result of the difference calculation is compared with the step size to obtain the temperature change rate. Preferably, the calculation step size is 60 seconds; The heat flow rate of the combustion chamber is calculated by multiplying the total heat capacity of the combustion chamber by the rate of temperature change. Based on the calculation logic of the combustion chamber heat flow rate, the heat flow rate of the flue gas passage and the heat exchange component are obtained respectively. The total wall heat flow rate is obtained by summing the heat flow rate of the combustion chamber, the heat flow rate of the flue gas passage and the heat exchange component. It should be noted that the total wall heat flow rate reflects the following physical meaning: it quantifies the dynamic heat exchange rate of heat storage / release generated by the thermal inertia of the combustion chamber, flue gas passage, and heat exchange component walls under extremely slow operating conditions of the gas turbine unit. It directly reflects the intensity and scale of the wall thermal hysteresis effect, and also clarifies the direction of heat transfer between the wall and the flue gas and working fluid during the slow increase and decrease of load: when the load increases, the total wall heat flow rate is positive, which means that the wall continuously absorbs heat from the flue gas and forms heat storage hysteresis; when the load decreases, the total wall heat flow rate is negative, which means that the wall continuously releases the accumulated heat and forms heat release hysteresis. Step 2: Calculate the theoretical steady-state combustion temperature based on the total wall heat flux using the steady-state thermodynamic balance method, and perform drift analysis to obtain the combustion temperature drift; then, according to the load conditions, superimpose the theoretical steady-state combustion temperature and the combustion temperature drift to obtain the actual combustion temperature. In step two, the theoretical steady-state combustion temperature is calculated using the steady-state thermodynamic balance method based on the total wall heat flux, and the combustion temperature drift is obtained through drift analysis as follows: The real-time load change rate, rated load of the gas turbine unit, standard intake air temperature, standard intake air pressure and design steady-state air-fuel ratio within the extremely slow-changing operating condition window are obtained as the basic calculation parameters. The theoretical steady-state combustion temperature under the current load is calculated using the steady-state thermodynamic balance method. It is understandable that the theoretical steady-state combustion temperature is the ideal combustion temperature when there is no thermal inertia hysteresis and no wall heat storage and release interference. As the benchmark value for this temperature drift correction, the calculation process only uses fixed parameters of unit design and real-time operating condition data, without empirical fitting and weight assignment. The real-time total flue gas volume is collected by the monitoring system within the window of extremely slow change operating conditions. The real-time total flue gas volume can reflect the total amount of flue gas emissions during the extremely slow change adjustment process. The heat transfer coefficient U1 and effective heat transfer area A1 of the combustion chamber, the heat transfer coefficient U2 and effective heat transfer area A2 of the flue gas passage, and the heat transfer coefficient U3 and effective heat transfer area A3 of the heat exchange components are obtained respectively. The comprehensive heat transfer coefficient is then calculated using the area-weighted average method from engineering thermophysics. Specifically: The heat transfer coefficient and effective heat exchange area are taken from the gas turbine unit design manual and are inherent physical parameters of the equipment. The overall heat transfer coefficient is obtained by dividing [(U1×A1)+(U2×A2)+(U3×A3)] / (A1+A2+A3); Obtain the specific heat capacity of flue gas at constant pressure; It is understandable that the flue gas specific heat capacity at constant pressure, like the heat transfer coefficient and the effective heat exchange area, are all taken from the gas turbine unit design manual and are inherent physical parameters of the equipment. The flue gas specific heat capacity at constant pressure refers to the amount of heat required to raise the temperature of a unit mass of flue gas by 1°C under the actual operating conditions of the unit with basically constant pressure. The combustion temperature drift is obtained by multiplying the real-time total flue gas volume, the comprehensive heat transfer coefficient and the flue gas specific heat capacity at constant pressure, and then comparing the total wall heat flow rate with the result of the product calculation. In step two, the process of superimposing the theoretical steady-state combustion temperature and the combustion temperature drift according to the load conditions to obtain the actual combustion temperature is as follows: Determine the load adjustment direction for the current extremely slow-changing operating condition based on the calculation step size set in step one; The monitoring system is used to obtain the real-time load of the gas turbine unit at the current data collection time and compare it with the past load of the gas turbine unit at the previous data collection time. If the real-time load is greater than the past load, it is determined to be a load increase condition. At this time, the combustion chamber, flue gas passage, and heat exchange component walls are in a heat storage state, absorbing heat from the combustion zone. The actual combustion temperature is lower than the theoretical steady-state temperature, and the combustion temperature drift is negative. If the real-time load is less than the past load, it is determined to be a load reduction condition. At this time, the wall is in a heat release state, releasing the previously stored heat to the combustion zone. The actual combustion temperature is higher than the theoretical steady-state temperature, and the combustion temperature drift is taken as a positive value. If the real-time load is equal to the past load, it is determined to be a steady-state transition condition. There is no obvious heat storage or heat release behavior on the wall, the combustion temperature drift is 0, and no temperature correction is performed. The actual combustion temperature is obtained by summing the theoretical steady-state combustion temperature and the combustion temperature drift after directional assignment. Understandably, this correction method directly offsets the temperature shift caused by the thermal inertia hysteresis of the wall from the heat source end, so that the combustion temperature no longer lags and can completely follow the AGC extremely slow load change command synchronously. Step 3: Obtain the real-time opening degree of the gas regulating valve and determine whether it is in the slightly open sensitive range; if so, construct an air-fuel ratio correction model based on the actual combustion temperature to obtain the actual air-fuel ratio, and perform mismatch analysis to obtain the theoretical gas flow rate to adjust the real-time opening degree of the gas regulating valve. In step three, the process of obtaining the real-time opening degree of the gas regulating valve and determining whether it is within the slightly open sensitive range is as follows: The gas regulating valve position feedback signal is collected in real time by the monitoring system, and the real-time opening degree of the gas regulating valve is obtained based on the valve position feedback signal. It should be noted that the valve position feedback signal is directly taken from the closed-loop feedback loop of the gas turbine control system. There is no signal conversion or delay distortion, and it can directly reflect the actual flow section ratio of the regulating valve, providing the original basis for determining the air-fuel ratio sensitive range. Compare the real-time opening of the gas regulating valve with the preset range; If the real-time opening degree is within the preset range, it is determined that the gas regulating valve has entered the micro-opening sensitive range. In this range, the flow characteristics of the regulating valve are strongly nonlinear. Even a small opening degree fluctuation can cause a large jump in gas flow. This is the core range of action caused by wall thermal hysteresis leading to dynamic mismatch of air-fuel ratio. After the determination, the air-fuel ratio correction module is immediately activated. If the real-time opening degree is not within the preset range, the flow characteristics of the regulating valve are close to linear, and the air-fuel ratio is extremely insensitive to the drift of combustion temperature. Therefore, the air-fuel ratio correction logic will not be executed temporarily to avoid ineffective regulation that could interfere with the stable operation of the unit. In step three, if the actual air-fuel ratio is obtained by constructing a corrected air-fuel ratio model based on the true combustion temperature, and performing mismatch analysis to obtain the theoretical gas flow rate, the process of adjusting the real-time opening of the gas regulating valve is as follows: The specific process of constructing the air-fuel ratio correction model is as follows: The current methane volume is obtained based on the real-time component parameters of the gas collected by the monitoring system. Among them, the real-time composition parameters of the gas represent the real-time quantitative data of the chemical composition and volume percentage of the gas in the gas turbine unit. Based on the chemical reaction formula for the complete combustion of methane, it can be concluded that the complete combustion of a unit volume of methane requires twice the volume of oxygen. Use twice the current methane volume as the theoretical oxygen volume; calculate the theoretical air volume by dividing the theoretical oxygen volume by 21% of the oxygen content in the air. The effect of temperature on air volume is corrected using the ideal gas law: The real-time theoretical air volume is calculated by multiplying the ratio of the actual combustion temperature to the standard thermodynamic temperature and then multiplying the result of the ratio with the theoretical air volume. Preferably, the standard thermodynamic temperature is 273.15 K; The real-time total air volume is obtained by using a monitoring system, and the actual air-fuel ratio is calculated by comparing the real-time total air volume with the real-time theoretical air volume. The design air-fuel ratio corresponding to the real-time load of the gas turbine unit is extracted based on the gas turbine unit design manual; the air-fuel ratio mismatch is calculated by the difference between the actual air-fuel ratio and the design air-fuel ratio. If the air-fuel ratio mismatch is positive, it means that the actual required air-fuel ratio at the current actual combustion temperature is higher than the steady-state value, and the gas supply is relatively excessive; if the air-fuel ratio mismatch is negative, it means that the actual required air-fuel ratio is lower than the steady-state value, and the gas supply is relatively insufficient. The ratio of the air-fuel ratio mismatch to the design air-fuel ratio is calculated, and the result of the ratio calculation is multiplied by the theoretical steady-state flow rate to obtain the gas compensation amount. Understandably, if the gas compensation is positive, the gas supply needs to be increased; if the gas compensation is negative, the gas supply needs to be reduced. The theoretical gas flow rate is obtained by summing the gas compensation amount and the instantaneous volumetric flow rate. The theoretical gas flow generation command is input into the gas turbine control system. The gas turbine control system adjusts the real-time opening of the gas regulating valve according to the target flow in a closed loop, so that the actual gas supply of the regulating valve is completely matched with the actual combustion temperature and the AGC extremely slow load change rate, so that the air-fuel ratio is always maintained in the optimal range of complete combustion. The technical solution of this invention is as follows: Identify the extremely slow-changing operating condition window of the gas turbine unit and construct a scatter plot; determine whether a closed narrow hysteresis loop occurs; if so, determine that a nonlinear hysteresis effect has occurred, and perform wall thermal inertia analysis to obtain the total wall heat flow rate; identify the nonlinear hysteresis effect, quantify the heat exchange caused by the wall thermal inertia, and provide a reliable quantitative basis for subsequent hysteresis correction; calculate the theoretical steady-state combustion temperature based on the total wall heat flow rate using the steady-state thermodynamic balance method, and perform drift analysis to obtain the combustion temperature drift; superimpose the load condition and the theoretical steady-state combustion temperature to obtain the actual combustion temperature; eliminate the combustion temperature drift caused by wall thermal inertia, allowing the actual combustion temperature to change synchronously with the AGC extremely slow-changing load command, and establish a hysteresis-free temperature reference; obtain the real-time opening degree of the gas regulating valve and determine whether it is in the slightly open sensitive range; if so, construct an air-fuel ratio correction model based on the actual combustion temperature to obtain the actual air-fuel ratio, perform mismatch analysis to obtain the theoretical gas flow rate, and adjust the real-time opening degree; avoid flow mismatch through dynamic air-fuel ratio correction, ensuring that the unit is always in the optimal combustion state.

[0023] Example 2 Please see Figure 1 As shown, a method for predicting the power consumption of a gas turbine unit includes: Step 4: Calculate the instantaneous power consumption based on the theoretical gas flow rate; retrieve the theoretical power consumption from the historical steady-state database of the gas turbine unit, set the hysteresis reference values ​​for load increase and decrease conditions respectively, and perform distortion analysis in combination with the instantaneous power consumption to obtain the actual gas consumption value; Obtain the current power generation efficiency at the current data collection time point, and calculate the instantaneous power consumption by comparing the theoretical gas flow rate with the current power generation efficiency; Instantaneous power consumption refers to the real-time gas consumption per unit of power generated by the gas turbine unit at the current data collection point. The instantaneous power consumption has completed the air-fuel ratio mismatch correction, leaving only a narrow hysteresis deviation caused by the nonlinear hysteresis of wall heat storage / release. In step four, the process of performing distortion analysis to obtain the true gas consumption value is as follows: Retrieve the historical steady-state database of the gas turbine unit, and match the theoretical power consumption corresponding to the current real-time load based on the historical steady-state database; It should be noted that the theoretical power consumption is the ideal gas consumption value with no thermal inertia hysteresis, no adjustment deviation, and no air-fuel ratio mismatch. Based on the real-time gas consumption difference sequence from step one and the theoretical power consumption for power generation, hysteresis baseline values ​​corresponding to load increase and load decrease conditions are constructed respectively: Load increase condition: Calculate the difference between the unit base value and the narrow offset threshold, and multiply the result of the difference calculation with the theoretical power consumption to obtain the hysteresis base value for load increase; Load reduction condition: The unit base value and the narrow offset threshold are accumulated and calculated. The result of the accumulated calculation is multiplied with the theoretical power consumption to obtain the hysteresis base value for load reduction. It should be noted that the unit base value is 1 and the narrow offset threshold is 0.15; the narrow offset threshold is the inherent narrow offset threshold of gas consumption caused by wall thermal inertia under extremely slowly changing operating conditions, which is directly quantified and determined by the hysteresis closed-loop interval of the real-time gas consumption difference sequence in step one. Following the load condition determination logic from step two, the load adjustment direction within the current extremely slow-changing condition window is verified in real time: Compare the real-time load at the current data collection time point with the past load at the previous data collection time point; If the real-time load is greater than the past load, it is determined to be a load increase condition, and the hysteresis reference value of the load increase is matched. If the real-time load is less than the past load, it is determined to be a load reduction condition, and the hysteresis reference value for load reduction is matched. If the real-time load is equal to the past load, it is determined to be a steady-state transition condition. There is no obvious heat storage or release behavior on the wall, and no hysteresis error is generated. The theoretical power consumption is directly used as the actual gas consumption. If the load is determined to be increasing, the absolute deviation value is obtained by calculating the absolute difference between the instantaneous power consumption and the hysteresis reference value of the load increase. If the condition is determined to be a load reduction condition, the absolute deviation of gas consumption is calculated by the absolute difference between the instantaneous generator power consumption and the hysteresis reference value of the load reduction condition. In some embodiments, the absolute deviation of gas consumption is compared with a narrow-amplitude hysteresis fluctuation determination threshold; Among them, the threshold for judging narrow-range hysteresis fluctuation is 0.1% of the theoretical power consumption; If the absolute deviation of gas consumption is less than the threshold for judging narrow-range hysteresis fluctuation, then the gas consumption deviation is judged to be a narrow-range hysteresis fluctuation caused by wall thermal inertia hysteresis, which is a distortion component that is not in real operation and an error elimination operation needs to be performed. Specifically, the error elimination operation is as follows: if it is a load increase condition, the hysteresis reference value of the load increase is marked as the actual gas consumption value at the current collection time point; if it is a load decrease condition, the hysteresis reference value of the load decrease is marked as the actual gas consumption value at the current collection time point. It should be noted that the error elimination operation directly removes the regular narrow-range deviation of low gas consumption when increasing load and high gas consumption when decreasing load, so that the instantaneous gas consumption is freed from the nonlinear interference of wall thermal inertia and returns to the true level that is completely matched with the real-time load, the actual combustion temperature, and the corrected air-fuel ratio. If the absolute deviation of gas consumption is greater than or equal to the threshold for judging narrow-range hysteresis fluctuation, the deviation is judged as a normal operating condition deviation such as unit load adjustment and gas composition fluctuation, which is unrelated to wall thermal hysteresis. Therefore, no rejection operation is performed, and the instantaneous power consumption is retained as the true gas consumption value. It is understandable that the physical meaning of the real gas consumption value is: the real-time gas consumption per unit power of the gas turbine unit under the current real-time load, real combustion temperature and optimal air-fuel ratio matching state, with no hysteresis interference. Step 5: Calculate the gas consumption at a single point based on the actual gas consumption value, and sum them up to obtain the total gas consumption at the hourly and daily levels; perform range calibration analysis based on the total gas consumption at the hourly and daily levels to obtain the final gas consumption prediction range for the hour and day. Based on the actual gas consumption value obtained in step four, the gas consumption at a single point is calculated by multiplying the actual gas consumption value at the time of collection, the actual power generation efficiency, and the time step. Preferably, the time step is 1 second; The gas consumption at a single point is continuously accumulated according to the extremely slow-changing operating conditions to obtain the total gas consumption at the hour level. Then, the total gas consumption at the hour level is used as the metering unit to continuously accumulate hourly to obtain the total gas consumption at the daily level. In step five, the process of performing range calibration analysis to obtain the hourly and daily final gas consumption prediction range is as follows: Retrieve the hourly and daily reference gas consumption values ​​predicted by the generator gas consumption prediction model under extremely slowly changing operating conditions; The hourly reference gas consumption value and the daily reference gas consumption value are multiplied by the system deviation range to obtain the hourly reference deviation range and the daily reference deviation range, respectively. It should be noted that the system deviation ranges from 0.3% to 0.6%; this deviation is formed by the long-term continuous accumulation of wall heat storage and release hysteresis, dynamic mismatch of air-fuel ratio, and narrow hysteresis loop error, and is unrelated to the real-time adjustment status of the unit. It belongs to a fixed systematic deviation that cannot be corrected by the steady-state gas consumption prediction model. The final gas consumption prediction range is obtained by multiplying the actual gas consumption value by the maximum and minimum values ​​of the time reference deviation range. The final daily gas consumption prediction range is obtained by multiplying the actual gas consumption value by the maximum and minimum values ​​of the daily baseline deviation range. It should be noted that the significance of the hourly final gas consumption prediction range and the daily final gas consumption prediction range is that they are both based on the actual gas consumption value and are accumulated point by point according to a uniform time step; they reflect the actual total gas consumption range predicted by the gas unit under the extremely slow change regulation condition of the grid deep peak shaving AGC, without dynamic hysteresis interference, regulation mismatch deviation, and error accumulation within a single hour or single day cycle. The technical solution of this invention is as follows: Instantaneous power consumption is calculated based on theoretical gas flow rate; theoretical power consumption from the historical steady-state database of the gas turbine unit is retrieved to set a hysteresis benchmark value, and distortion analysis is performed in conjunction with instantaneous power consumption to obtain the true gas consumption value; the narrow hysteresis distortion component caused by wall thermal inertia is effectively removed to obtain pure true gas consumption without hysteresis interference, which truly reflects the gas consumption level of the unit itself; single-point gas consumption is calculated based on the true gas consumption value, and accumulated to obtain hourly and daily total gas consumption; range calibration analysis is performed to obtain the final hourly and daily gas consumption prediction range; the hourly and daily total gas consumption are accumulated, and a gas consumption prediction range without dynamic hysteresis and error accumulation is given, which greatly improves the accuracy and practicality of gas consumption prediction under extremely slow-changing operating conditions.

[0024] Example 3 Please see Figure 3 As shown, a predictive system for the power consumption of a gas turbine unit includes the following modules: Hysteresis identification module: Identifies the extremely slow-changing operating condition window of the gas turbine unit and constructs a scatter plot; determines whether a closed narrow hysteresis loop occurs. If it does, it determines that a nonlinear hysteresis effect has occurred and performs wall thermal inertia analysis to obtain the total wall heat flux. Temperature correction module: The theoretical steady-state combustion temperature is calculated based on the total wall heat flux using the steady-state thermodynamic balance method, and the combustion temperature drift is obtained by drift analysis; the actual combustion temperature is obtained by superimposing the load condition and the theoretical steady-state combustion temperature. Opening adjustment module: acquires the real-time opening of the gas regulating valve, determines whether it is in the slightly open sensitive range; if so, it constructs an air-fuel ratio correction model based on the actual combustion temperature to obtain the actual air-fuel ratio, performs mismatch analysis to obtain the theoretical gas flow rate, and adjusts the real-time opening. Gas consumption analysis module: Calculates instantaneous power consumption based on theoretical gas flow rate; retrieves theoretical power consumption from historical steady-state database of gas turbine unit to set hysteresis benchmark value, and performs distortion analysis in conjunction with instantaneous power consumption to obtain the actual gas consumption value; Prediction output module: Calculates the gas consumption at a single point based on the actual gas consumption value, accumulates it to obtain the total gas consumption at the hourly and daily levels; and performs range calibration analysis to obtain the final gas consumption prediction range for the hour and day.

[0025] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A method for predicting the power consumption of a gas turbine unit, characterized in that: Identify the extremely slow-changing operating condition window of the gas turbine unit and construct a scatter plot; determine whether a closed narrow hysteresis loop occurs. If it does, determine that a nonlinear hysteresis effect has occurred, and perform wall thermal inertia analysis to obtain the total wall heat flux. The theoretical steady-state combustion temperature is calculated using the steady-state thermodynamic balance method based on the total wall heat flux, and the combustion temperature drift is obtained by drift analysis. The actual combustion temperature is then obtained by superimposing the load conditions and the theoretical steady-state combustion temperature. Obtain the real-time opening degree of the gas regulating valve and determine whether it is in the slightly open sensitive range; If it is in the true combustion temperature, an air-fuel ratio correction model is constructed to obtain the actual air-fuel ratio, a mismatch analysis is performed to obtain the theoretical gas flow rate, and the real-time opening degree is adjusted. Instantaneous power consumption is calculated based on theoretical gas flow rate; theoretical power consumption from historical steady-state database of gas turbine unit is retrieved to set hysteresis reference value, and distortion analysis is performed in conjunction with instantaneous power consumption to obtain actual gas consumption value; The gas consumption at a single point is calculated based on the actual gas consumption value, and then accumulated to obtain the total gas consumption at the hourly and daily levels. The range calibration analysis was then performed to obtain the hourly and daily final gas consumption prediction range.

2. The method for predicting the power consumption of a gas turbine unit according to claim 1, characterized in that: The process of constructing a scatter plot involves: The actual power generation efficiency and instantaneous volumetric flow rate of the gas turbine unit at different data collection time points within the extremely slow-changing operating condition window are integrated to obtain the actual power generation efficiency sequence and the instantaneous volumetric flow rate sequence. Obtain the theoretical steady-state flow rate sequence, and calculate the difference between the instantaneous volumetric flow rate sequence and the theoretical steady-state flow rate sequence to obtain the real-time gas consumption difference; A scatter plot was constructed with the actual power generation efficiency at the time of data collection as the horizontal axis and the real-time gas consumption difference as the vertical axis.

3. The method for predicting the power consumption of a gas turbine unit according to claim 1, characterized in that: The process of determining the occurrence of nonlinear hysteresis effect is as follows: In the scatter plot, if the real-time gas consumption difference corresponding to the actual power generation efficiency during the load increase process is below the horizontal axis, and the real-time gas consumption difference corresponding to the actual power generation efficiency during the load decrease process is above the horizontal axis, then a closed narrow hysteresis loop is determined to have occurred. If a closed narrow hysteresis loop appears, it is determined that the gas turbine unit is experiencing a nonlinear hysteresis effect within the extremely slowly changing operating condition window.

4. The method for predicting the power consumption of a gas turbine unit according to claim 1, characterized in that: The process of obtaining the combustion temperature drift by performing drift analysis is as follows: The total flue gas volume, comprehensive heat transfer coefficient, and flue gas specific heat capacity at constant pressure are obtained in real time and multiplied to calculate the total wall heat flow rate. The ratio of the total wall heat flow rate to the product calculation result is then used to obtain the combustion temperature drift.

5. The method for predicting the power consumption of a gas turbine unit according to claim 4, characterized in that: The process of obtaining the overall heat transfer coefficient is as follows: Based on the gas turbine design manual, the heat transfer coefficient U1 and effective heat exchange area A1 of the combustion chamber, the heat transfer coefficient U2 and effective heat exchange area A2 of the flue gas passage, and the heat transfer coefficient U3 and effective heat exchange area A3 of the heat exchange component were obtained respectively. The comprehensive heat transfer coefficient is calculated by combining the area-weighted average method in engineering thermophysics.

6. The method for predicting the power consumption of a gas turbine unit according to claim 1, characterized in that: The process of performing mismatch analysis is as follows: The air-fuel ratio mismatch is calculated by obtaining the difference between the actual air-fuel ratio and the design air-fuel ratio. Obtain the theoretical steady-state flow rate, calculate the ratio of the air-fuel ratio mismatch to the design air-fuel ratio, and multiply the ratio calculation result with the theoretical steady-state flow rate to calculate the gas compensation amount; The corrected theoretical gas flow rate is obtained by summing the gas compensation amount and the instantaneous volumetric flow rate. The theoretical gas flow rate generation command is input into the gas turbine control system, which then adjusts the real-time opening of the gas regulating valve according to the target flow rate in a closed-loop manner.

7. The method for predicting the power consumption of a gas turbine unit according to claim 6, characterized in that: The actual air-fuel ratio is obtained as follows: The current methane volume is obtained based on the real-time component parameters of the gas collected by the monitoring system. The real-time theoretical air volume is obtained by analyzing and calculating the current methane volume and the chemical reaction formula for the complete combustion of methane. The real-time total air volume is obtained by using a monitoring system, and the actual air-fuel ratio is calculated by comparing the real-time total air volume with the real-time theoretical air volume.

8. The method for predicting the power consumption of a gas turbine unit according to claim 1, characterized in that: The process of setting the hysteresis reference value for load increase and decrease conditions is as follows: Obtain theoretical power consumption, unit base value, and narrow-range offset threshold; The difference between the unit base value and the narrow-range offset threshold is calculated, and the result of the difference calculation is multiplied by the theoretical power consumption of the generator to obtain the hysteresis base value for load increase. The unit base value and the narrow-range offset threshold are accumulated and calculated. The result of the accumulated calculation is multiplied with the theoretical power consumption to obtain the hysteresis base value for load reduction.

9. The method for predicting the power consumption of a gas turbine unit according to claim 1, characterized in that: The process of performing range calibration analysis is as follows: The time-based and daily-based reference gas consumption values ​​predicted by the steady-state gas consumption prediction model under extremely slowly changing operating conditions are retrieved and the deviation range is calculated to obtain the time-based and daily-based deviation ranges. The final gas consumption prediction range for the hour / day is calculated by multiplying the actual gas consumption value by the maximum and minimum values ​​of the hourly / dayly reference deviation range.

10. A predictive system for the power consumption of a gas turbine unit according to claim 1, used to implement the predictive method for the power consumption of a gas turbine unit according to any one of claims 1-9, characterized in that: Including the following: Hysteresis identification module: Identifies the extremely slow-changing operating condition window of the gas turbine unit and constructs a scatter plot; determines whether a closed narrow hysteresis loop occurs. If it does, it determines that a nonlinear hysteresis effect has occurred and performs wall thermal inertia analysis to obtain the total wall heat flux. Temperature correction module: The theoretical steady-state combustion temperature is calculated based on the total wall heat flux using the steady-state thermodynamic balance method, and the combustion temperature drift is obtained by drift analysis; and the actual combustion temperature is obtained by superimposing the load conditions and the theoretical steady-state combustion temperature. Opening adjustment module: acquires the real-time opening of the gas regulating valve and determines whether it is in the slightly open sensitive range; If it is in the true combustion temperature, an air-fuel ratio correction model is constructed to obtain the actual air-fuel ratio, a mismatch analysis is performed to obtain the theoretical gas flow rate, and the real-time opening degree is adjusted. Gas consumption analysis module: Calculates instantaneous power consumption based on theoretical gas flow rate; retrieves theoretical power consumption from historical steady-state database of gas turbine unit to set hysteresis benchmark value, and performs distortion analysis in conjunction with instantaneous power consumption to obtain the actual gas consumption value; Prediction output module: Calculates the gas consumption at a single point based on the actual gas consumption value, accumulates it to obtain the total gas consumption at the hourly and daily levels; and performs range calibration analysis to obtain the final gas consumption prediction range for the hour and day.

Images