Supercritical co2 power generation system adapted to fluctuating industrial waste heat and control method thereof

Abstract

The application provides a supercritical CO2 power generation system suitable for adapting to industrial waste heat with severe fluctuation and a control method thereof, and relates to the technical field of supercritical CO2 power generation.The application makes CO2 pass through compression, heat recovery, injection mixing, heating, expansion work and cooling in sequence through a closed cycle loop, and then return to a low-pressure CO2 storage tank; the fluctuation of a heat source is buffered and compensated through a heat storage tank, a hot water inlet valve group, a steam-water mixed heater, a flash valve group and a flash steam supplement device; the cycle pressure is compensated through a high-pressure supercritical CO2 storage tank, a pressure buffer container and a high-pressure injection mixer. The control method forms operation criteria, heat source power prediction results, heat source adjustment targets and working medium control targets according to operation states, power generation control information and historical operation sequences, solves control amounts in combination with safety constraints, and outputs linkage control instructions. The application is beneficial to improving the operation stability and energy utilization efficiency of the supercritical CO2 power generation system under the condition of industrial waste heat with severe fluctuation.

Description

Technical Field

[0001] This invention relates to the field of supercritical CO2 power generation technology, and particularly to a supercritical CO2 power generation system and its control method adapted to drastically fluctuating industrial waste heat. Background Technology

[0002] Supercritical CO2 power generation technology, characterized by its compact system, high heat exchange efficiency, and suitability for utilizing medium- and low-temperature heat sources, has become an important development direction in the field of industrial waste heat recovery. In typical process industrial production processes, waste heat emissions are often located in the medium- and low-temperature range, and are affected by process load switching, intermittent emissions, and changes in production cycle, resulting in significant random fluctuations in heat source power, with a large difference between instantaneous peak values ​​and average levels.

[0003] Existing supercritical CO2 power generation systems are mostly designed around stable heat sources and rated operating conditions. When faced with large fluctuations in waste heat, the hot-end buffer response is insufficient, the cycle parameters cannot be matched with changes in the heat source in a timely manner, and the cold-end regulation and energy distribution cannot balance operating efficiency and economic benefits. When the heat source rises or falls suddenly or the load changes abruptly, the system may also experience operational risks such as low-flow instability of the compressor equipment, overheating of the heat exchange outlet, and untimely energy release on the high-pressure side.

[0004] Furthermore, existing control methods typically treat heat source fluctuations as deterministic disturbances, making it difficult to predict and manage the uncertainty of waste heat power, which can easily lead to a conflict between safety margin and power generation revenue. Therefore, there is an urgent need for a supercritical CO2 power generation technology that can adapt to drastically fluctuating industrial waste heat while balancing efficiency, economy, and operational safety. Summary of the Invention

[0005] To address the problems of difficulty in mitigating heat source shocks during the recovery of industrial waste heat with severe fluctuations, efficiency decline after the supercritical CO2 cycle deviates from its rated operating conditions, lag in power generation system pressure regulation, and insufficient safety margin under extreme fluctuation conditions, this invention provides a supercritical CO2 power generation system adapted to industrial waste heat with severe fluctuations and its control method.

[0006] To address the aforementioned problems, this invention provides a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat, comprising a closed-loop circulation system, a heat source buffer branch, a pressure compensation branch, and a controller. In the closed-loop circulation system, the outlet of the low-pressure CO2 storage tank is connected to the injection inlet of the high-pressure jet mixer via a CO2 compression pump unit, a low-temperature working fluid channel of a regenerating heat exchanger, and a check valve. The outlet of the high-pressure jet mixer is connected back to the low-pressure CO2 storage tank via a high-pressure CO2 heater, a turbine inlet regulating valve, a turbine, a high-temperature working fluid channel of a regenerating heat exchanger, and a cooling unit. In the heat source buffer branch, a hot water inlet valve group is connected to a heat storage tank. The hot water outlet of the heat storage tank is connected to the hot water inlet of the high-pressure CO2 heater via a steam-water mixing heater. The flash steam replenishment device is equipped with a flash valve group, and its steam outlet is connected to the steam inlet of the steam-water mixing heater. In the pressure compensation branch, the high-pressure supercritical CO2 storage tank, whose operating pressure is higher than the target pressure at the outlet of the high-pressure jet mixer, is connected to the working fluid inlet of the high-pressure jet mixer via a pressure regulating valve, a pressure buffer container, and an injection valve. The controller is connected to the CO2 compressor pump set, turbine inlet regulating valve, pressure regulating valve, jet valve, hot water inlet valve set, and flash evaporation valve set via signal connection.

[0007] Furthermore, the thermal storage tank is divided into a first thermal storage zone and a second thermal storage zone, with a hot water inlet valve group connecting the two zones. The first thermal storage zone contains an encapsulated phase change thermal storage unit and a shell-side saturated water space, with a flash steam replenishment device drawing water from the shell-side saturated water space via a flash valve group. The second thermal storage zone contains a hot water storage space and a heat replenishment interface. Thus, the first thermal storage zone can absorb short-term thermal shocks through phase change thermal storage and compensate for sudden drops in heat source through shell-side saturated water flash steam replenishment, while the second thermal storage zone can provide a buffer of medium-temperature sensible heat capacity.

[0008] Furthermore, the power generation system also includes a hot-water lithium bromide chiller, a chilled water storage tank, a cold storage distribution three-way valve, a subcooling heat exchanger, and auxiliary cooling branches. The hot water inlet of the hot-water lithium bromide chiller and the hot water inlet of the high-pressure CO2 heater are both connected to hot water supply pipelines led from a steam-water mixing heater. The chilled water outlet of the hot-water lithium bromide chiller is connected to the chilled water storage tank. The chilled water storage tank is connected to the subcooling heat exchanger and the auxiliary cooling branches via the cold storage distribution three-way valve. The subcooling heat exchanger is connected in series between the cooling unit outlet and the low-pressure CO2 storage tank inlet. The auxiliary cooling branches are connected to the turbine oil cooler and the generator cooling water jacket. The cold storage distribution three-way valve is connected to the controller signal.

[0009] Furthermore, the power generation system also includes a compressor pump return pipe, a cold flow bypass, and an emergency pressure reduction bypass. The compressor pump return pipe is connected between the outlet of the CO2 compressor pump unit and the gas phase zone of the low-pressure CO2 storage tank and is equipped with a first fast-acting regulating valve. The cold flow bypass is connected between the inlet and outlet of the high-pressure CO2 heater and is equipped with a second fast-acting regulating valve. The emergency pressure reduction bypass is connected between the pipe section before the turbine inlet regulating valve and the turbine outlet pipe section and is equipped with a quick-opening solenoid valve and a multi-stage throttling orifice plate. The first fast-acting regulating valve, the second fast-acting regulating valve, and the quick-opening solenoid valve are all connected to the controller signal.

[0010] The present invention also provides a control method for the above-mentioned supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat, comprising:

[0011] S101: Based on the operating status of the closed-loop circuit, heat source buffer branch, and pressure compensation branch, combined with power generation control information and historical operating sequences, identify the operating mode and form an operating benchmark.

[0012] S102: Based on the power generation control information and historical operating sequence, the heat source power prediction result is generated. The difference between the current heat source power and the operating benchmark is used to generate the heat source power change amount, and the heat source regulation target for regulating the heat storage tank and flash steam injection device is generated.

[0013] S103: Based on the available heat source power and ambient temperature, the working fluid control target is formed, which includes the turbine inlet pressure target and the CO2 mass flow rate target.

[0014] S104: Safety constraints are formed by the state of the circulating working fluid and equipment limits. The control quantity is solved by combining the heat source power prediction results, heat source adjustment target and working fluid control target, and the linkage control command is output to the CO2 compression pump group, turbine inlet regulating valve, pressure regulating valve, jet valve, hot water inlet valve group and flash evaporation valve group.

[0015] Further, in step S101, when the start command and heat source availability signal are valid, the CO2 compression pump group is in an unstarted state, and the main circulation pressure has not reached the start pressure, the operating mode is determined to be the black start mode; in step S104, in the black start mode, the pressure regulating valve and the injection valve are opened, so that the high-pressure supercritical CO2 storage tank supplies pressure to the working fluid inlet of the high-pressure CO2 heater through the pressure buffer container and the high-pressure injection mixer; after the state of the circulating working fluid meets the start conditions, the CO2 compression pump group is started to establish the main circulation, so that the high-pressure injection mixer enters the ejector mixing state.

[0016] Further, in step S102, the heat source history, meteorological and process characteristics are input into a Gaussian process regression model containing a squared exponential kernel, a periodic kernel and a white noise kernel to obtain the heat source power prediction mean and the heat source power prediction variance; when the prediction error exceeds the prediction error threshold or the prediction interval coverage is lower than the preset coverage lower limit, the Gaussian process regression model is updated to regenerate the heat source power prediction result.

[0017] Further, in step S102, the heat source regulation target includes a heat storage allocation target and a flash steam replenishment target. The heat storage allocation target is used to adjust the inflow ratio of the hot water inlet valve group to the first and second heat storage zones in the heat storage tank. The flash steam replenishment target is used to adjust the flash valve group to draw water from the shell-side saturated water space of the first heat storage zone and to replenish steam to the steam-water mixing heater via the flash steam replenishment device. When the heat source power change is positive and the ratio of the heat source power change to the control cycle duration exceeds the positive threshold, or the cumulative positive change within the sliding window exceeds the cumulative threshold, a heat storage allocation target that increases the inflow ratio to the first heat storage zone is generated. When the heat source power change is negative and the ratio of its absolute value to the control cycle duration exceeds the decrease threshold, and the outlet water temperature of the steam-water mixing heater is lower than the flash steam replenishment trigger temperature threshold corrected by the temperature decrease rate of the first heat storage zone, a flash steam replenishment target is generated. In step S104, the heat storage allocation target and the flash steam replenishment target are used as inputs to solve for the control quantity to determine the control quantity of the hot water inlet valve group or the flash valve group.

[0018] Further, in step S103, based on a closed-loop thermodynamic model including a CO2 compression pump unit, a regenerative heat exchanger, a high-pressure jet mixer, a high-pressure CO2 heater, a turbine, and a cooling unit, with the goal of maximizing cycle efficiency, the correspondence between available heat source power, ambient temperature, turbine inlet pressure, and CO2 mass flow rate is solved offline under equipment constraints to construct a pressure-flow mapping table; based on the deviation between the measured cycle efficiency and the predicted cycle efficiency, the recursive least squares method or the extreme gradient boosting algorithm is used to correct the neighborhood mapping nodes of the operating condition.

[0019] Further, in step S104, the predicted mean of heat source power is input into a thermodynamic model including a heat storage tank, a steam-water mixing heater, a high-pressure CO2 heater, a high-pressure supercritical CO2 storage tank, a pressure buffer vessel, a high-pressure jet mixer, and a turbine for forward calculation. The predicted variance of heat source power is then transferred to the turbine inlet temperature and the high-pressure supercritical CO2 storage tank pressure according to the sensitivity or unscented transformation of the thermodynamic model at the current operating point, to obtain the predicted mean and standard deviation of the turbine inlet temperature and the high-pressure supercritical CO2 storage tank pressure. The predicted mean and standard deviation are then converted into deterministic boundary constraints according to a preset confidence level. In each control cycle, the control quantity sequence in the prediction time domain is solved, and the first control quantity in the control quantity sequence is output.

[0020] Compared with the prior art, the present invention has at least the following beneficial effects:

[0021] By coordinating the heat storage tank, steam-water mixing heater, hot water inlet valve group, flash valve group and flash steam replenishment device, the heat source can be shaped to handle sudden increases and decreases in industrial waste heat power, reducing the amplitude of heat source fluctuations directly transmitted to the supercritical CO2 cycle.

[0022] By providing pressure compensation to the closed-loop system through high-pressure supercritical CO2 storage tanks, pressure buffer vessels, and high-pressure jet mixers, the deviation of turbine inlet pressure and CO2 mass flow rate caused by heat source fluctuations can be reduced.

[0023] By using a hot water lithium bromide chiller unit, chilled water storage tank, cold storage distribution three-way valve, subcooling heat exchanger and auxiliary cooling branch, the system can distribute power between hot water supply, chilled water cold storage and CO2 subcooling, thereby improving the cold end regulation capability.

[0024] By using the compressor return pipe, cold flow bypass, and emergency pressure reduction bypass, bypass protection can be provided for low-load surge, hot-end over-temperature, and emergency shutdown over-pressure risks, thereby improving operational safety under drastic fluctuation conditions.

[0025] By using pressure-flow mapping tables and operating condition neighborhood mapping nodes for correction, turbine inlet pressure and CO2 mass flow rate can be adjusted according to available heat source power and ambient temperature.

[0026] By transferring the predicted variance of heat source power to the system state variables and forming deterministic boundary constraints, control quantities can be solved under the condition of considering random fluctuations in heat source, which is beneficial to improving the operational stability and energy utilization efficiency of supercritical CO2 power generation systems under conditions of drastically fluctuating industrial waste heat. Attached Figure Description

[0027] Figure 1 Overall process flow diagram of a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat.

[0028] Figure 2 This is a flowchart of the operation control of the heat source buffer.

[0029] Figure 3 This is a partial flowchart of a variable-condition adaptive dual-pressure power generation system.

[0030] Figure 4 This is the logic diagram of the decoupled control system for cold storage and power generation.

[0031] Figure 5 This is a partial P&ID diagram of a dual-redundant bridging system.

[0032] Figure 6 The flowchart shows the three-layer architecture of the stochastic model prediction and control integrated decision-making algorithm.

[0033] Among them, 1-turbine; 2-generator; 3-regenerative heat exchanger; 4-cooling unit; 5-hot water lithium bromide refrigeration unit; 6-low-pressure CO2 storage tank; 7-CO2 compression pump set; 8-high-pressure CO2 heater; 9-high-pressure supercritical CO2 storage tank; 10-heat storage tank; 11-thermal stratification baffle; 12-first heat storage zone; 13-second heat storage zone; 14-flash steam replenishment device; 15-hot water inlet valve group; 20-pressure buffer container; 21-high-pressure jet mixer; 25-chilled water storage tank; 26-cold storage distribution three-way valve; 27-subcooling heat exchanger; 29-turbine inlet regulating valve; 30-first quick regulating valve; 31-second quick regulating valve; 32-quick-opening solenoid valve; 33-multi-stage throttling orifice plate; 34-main circuit check valve; 35-pressure regulating valve; 36-jet valve. Detailed Implementation

[0034] This embodiment provides a supercritical CO2 power generation system and its control method adapted to drastically fluctuating industrial waste heat, applicable to low-temperature industrial waste heat recovery and power generation scenarios ranging from 80℃ to 150℃ in chemical and steel production processes. In this scenario, the industrial waste heat power can change significantly in a short period of time, for example, fluctuating within the range of 2MW to 20MW. This embodiment uses a heat source buffer branch to shape the industrial waste heat, a pressure compensation branch to compensate the pressure of the closed-loop circuit, a cold-end auxiliary regulation through a cold-end cold storage structure, and a controller to output linkage control commands based on prediction results, safety constraints, and revenue judgments.

[0035] like Figure 1 As shown, a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat includes a closed-loop circuit, a heat source buffer branch, a pressure compensation branch, a cold-end cold storage structure, a redundant crossover protection structure, and a controller. The closed-loop circuit includes a turbine 1, a generator 2, a regenerative heat exchanger 3, a cooling unit 4, a low-pressure CO2 storage tank 6, a CO2 compression pump set 7, a high-pressure CO2 heater 8, and a turbine inlet regulating valve 29; the heat source buffer branch includes a heat storage tank 10, a steam-water mixing heater, a flash steam replenishment device 14, and a hot water inlet valve set 15; the pressure compensation branch includes a high-pressure supercritical CO2 storage tank 9, a pressure regulating valve 35, a pressure buffer container 20, an injection valve 36, and a high-pressure injection mixer 21; the cold-end cold storage structure includes a hot-water lithium bromide chiller unit 5, a chilled water storage tank 25, a cold storage distribution three-way valve 26, a subcooling heat exchanger 27, and an auxiliary cooling branch; the redundant crossover protection structure includes a compressor return pipe, a cold flow bypass, and an emergency pressure reduction bypass.

[0036] In the closed-loop circuit, the outlet of the low-pressure CO2 storage tank 6 is connected to the injection inlet of the high-pressure jet mixer 21 via the CO2 compression pump set 7, the low-temperature working fluid channel of the regenerating heat exchanger 3, and the main check valve 34. The outlet of the high-pressure jet mixer 21 is connected back to the low-pressure CO2 storage tank 6 via the high-pressure CO2 heater 8, the turbine inlet regulating valve 29, the turbine 1, the high-temperature working fluid channel of the regenerating heat exchanger 3, and the cooling unit 4. The low-pressure CO2 storage tank 6 is used to temporarily store the low-pressure working fluid. The CO2 compression pump set 7 is used to pressurize the low-pressure CO2. The regenerating heat exchanger 3 is used to preheat the low-temperature working fluid after compression by the high-temperature working fluid discharged from the turbine 1. The high-pressure CO2 heater 8 is used to heat the CO2 using hot water supplied by the heat source buffer branch. The turbine 1 is used to expand and do work, and drive the generator 2 to generate electricity. The main line check valve 34 is located between the outlet of the low-temperature working medium channel of the regenerating heat exchanger 3 and the injection inlet of the high-pressure jet mixer 21. It is used to allow the main line working medium to enter the high-pressure jet mixer 21 in the forward direction and to prevent the high-pressure working medium of the pressure compensation branch from flowing back into the low-temperature working medium channel of the regenerating heat exchanger 3.

[0037] In the heat source buffer branch, the hot water inlet valve group 15 is connected to the heat storage tank 10. The hot water outlet of the heat storage tank 10 is connected to the hot water inlet of the high-pressure CO2 heater 8 via a steam-water mixing heater. The flash steam replenishment device 14 is equipped with a flash valve group, and its steam outlet is connected to the steam inlet of the steam-water mixing heater. The saturated steam generated by the flash steam replenishment device 14 enters the steam-water mixing heater and directly exchanges heat with the hot water and condenses, thus compensating for the temperature of the hot water entering the high-pressure CO2 heater 8 when the heat source drops suddenly. In the pressure compensation branch, the high-pressure supercritical CO2 storage tank 9 is connected to the working fluid inlet of the high-pressure jet mixer 21 via a pressure regulating valve 35, a pressure buffer container 20, and a jet valve 36. The high-pressure supercritical CO2 storage tank 9 can be a tank with a design pressure of 28 MPa and a rated operating pressure of 25 MPa. Under rated full load conditions, the target outlet pressure of the high-pressure jet mixer 21 can be set to 14 MPa. During sliding pressure operation, the target outlet pressure is adjusted by the pressure-flow mapping table according to the available heat source power.

[0038] The controller is connected to the CO2 compressor pump group 7 and controllable valves via signal connections. The controllable valves include a turbine inlet regulating valve 29, a pressure regulating valve 35, an injection valve 36, a hot water inlet valve group 15, a flash evaporation valve group, a cold storage distribution three-way valve 26, a first quick regulating valve 30, a second quick regulating valve 31, and a quick-opening solenoid valve 32. The first quick regulating valve 30, the second quick regulating valve 31, and the quick-opening solenoid valve 32 are all connected to the controller via signal connections. The controller also receives data on turbine inlet temperature, turbine inlet pressure, pressure of the high-pressure supercritical CO2 storage tank 9, CO2 mass flow rate, operating status of the CO2 compressor pump group 7, outlet temperature of the heat storage tank 10, outlet water temperature of the steam-water mixing heater, CO2 temperature at the outlet of the cooling unit 4, electricity price information, and upstream production process status. The circulating working fluid status includes the CO2 pressure, temperature, mass flow rate, pump speed, and opening degree of the turbine inlet regulating valve 29 within the closed-loop circulation system.

[0039] The heat storage tank 10 is divided into a first heat storage zone 12 and a second heat storage zone 13 by at least one thermal stratification partition 11, and is equipped with a hot water inlet valve group 15 connecting the first heat storage zone 12 and the second heat storage zone 13. The first heat storage zone 12 adopts a composite isolation structure of an encapsulated phase change heat storage unit and a shell-side saturated water space. The encapsulated phase change heat storage unit is made of stainless steel or aluminum encapsulation body, and the encapsulation body is filled with a phase change material with a phase change temperature near the target heating temperature, such as paraffin-based or hydrated salt-based composite PCM with a phase change temperature of about 112°C, and a heat storage density of not less than 180 kJ / kg. The outside of the encapsulation body is the shell-side saturated water space, which limits the working medium participating in flash evaporation to shell-side saturated water, thus spatially isolating the phase change material and the flash evaporation working medium. The flash evaporation steam replenishment device 14 draws water from the shell-side saturated water space through the flash evaporation valve group and performs depressurization flash evaporation. The second heat storage zone 13 is equipped with a hot water storage space and a heat replenishment interface. The heat replenishment interface is connected to a solar thermal collector system or an external auxiliary heat source to store medium-temperature hot water at 60°C to 90°C as a capacity buffer.

[0040] like Figure 2As shown, the heat source buffer branch switches its operating state according to the direction and magnitude of the heat source power change. When the industrial waste heat power increases, the controller judges the thermal shock intensity based on the ratio of the heat source power change to the control cycle duration and the cumulative positive change within the sliding window. When any judgment result reaches the corresponding trigger condition, the controller controls the hot water inlet valve group 15 to increase the inflow ratio of the first heat storage zone 12, allowing high-temperature hot water to preferentially enter the first heat storage zone 12. The encapsulated phase change heat storage unit absorbs heat and undergoes a solid-liquid phase change. When the industrial waste heat power decreases, the controller judges the steam replenishment demand based on the ratio of the absolute value of the heat source power change to the control cycle duration, the temperature drop rate of the first heat storage zone 12, and the outlet water temperature of the steam-water mixing heater. When the outlet water temperature of the steam-water mixing heater is lower than the flash steam replenishment trigger temperature threshold corrected by the temperature drop rate of the first heat storage zone 12, the controller opens the flash valve group of the flash steam replenishment device 14. The saturated water in the shell side of the first heat storage zone 12 is depressurized to form saturated steam, which is then replenished into the steam-water mixing heater. When industrial waste heat is in a low-power state for a long time, the controller can open the heat replenishment interface of the second heat storage zone 13 to maintain the temperature of the hot water storage space in the second heat storage zone 13 by the solar thermal system or external auxiliary heat source.

[0041] The total heat storage capacity of the thermal storage tank 10 can be calculated as follows:

[0042] ;

[0043] ;

[0044] ;

[0045] in, The total heat storage capacity of the heat storage tank 10 is... This refers to the heat storage capacity corresponding to the phase change material within the first heat storage zone 12. This refers to the sensible heat storage capacity corresponding to the hot water in the second heat storage zone 13. The mass of the phase change material in the encapsulated phase change thermal storage unit. The specific heat capacity of a solid-state phase change material. For the temperature rise of the solid phase change material before the phase change, For latent heat of phase transition, The specific heat capacity of liquid phase change materials, This refers to the temperature rise of the liquid phase change material after the phase change. For the hot water quality of the second thermal storage zone 13, The specific heat capacity of water, This refers to the temperature variation range of the hot water in the second heat storage zone 13. These parameters can be determined by the design volume of the heat storage tank 10, the mass of the phase change material filling, the amount of hot water filled, and the thermophysical property data of the material.

[0046] The steam replenishment power of the flash steam replenishment device 14 can be calculated as follows:

[0047] ;

[0048] in, For flash steam replenishment power, This refers to the flash steam mass flow rate. This represents the enthalpy of saturated steam at flash pressure. This represents the enthalpy of saturated water before flash evaporation. The pressure can be determined by the opening degree of the flash valve assembly, the pressure before the valve, the pressure after the valve, and the pressure in the gas-liquid separation space. and It can be determined according to the state tables of saturated water and saturated steam.

[0049] After acquiring operating status, power generation control information, and historical operating sequences, the controller forms an operating baseline. Operating status includes turbine inlet temperature, turbine inlet pressure, pressure of the high-pressure supercritical CO2 storage tank 9, CO2 mass flow rate, operating status of the CO2 compressor pump group 7, outlet temperature of the thermal storage tank 10, outlet water temperature of the steam-water mixing heater, CO2 temperature at the outlet of the cooling unit 4, and the opening degree of each valve. Power generation control information includes start-up commands, heat source availability signals, power generation priority signals, electricity price information, and load demand. Historical operating sequences include heat source power, heat source change rate, ambient temperature, process load, and system status data over a past period. The operating baseline includes the heat source power baseline, pressure baseline, temperature baseline, flow rate baseline, and safety boundary for the current control cycle.

[0050] When the controller determines, based on the power generation control information, that it has received a start-up command and a heat source availability signal, and that the CO2 compressor pump group 7 is not started and the main circulation pressure has not reached the start-up pressure, the controller sets the operating mode to black start mode. In the initial stage of black start mode, since the CO2 compressor pump group 7 has not yet started, the main circulation in the closed-loop circuit has not yet established an initial flow rate. The controller opens the pressure regulating valve 35 and the injection valve 36, allowing the working fluid in the high-pressure supercritical CO2 storage tank 9 to be injected into the high-pressure injection mixer 21 via the pressure buffer container 20. During this stage, the high-pressure injection mixer 21 does not actually eject fluid, but rather acts as a throttling and pressurizing channel to supply pressure to the downstream dead zone volume via a rapid response pressure source. After the circulating working fluid condition meets the start-up conditions, the controller starts the CO2 compressor pump group 7 to establish the main circulation. At this time, the injection inlet of the high-pressure injection mixer 21 establishes a clear pressure boundary and flow rate, and the system enters the injection mixing state. The start-up conditions include the main circulation pressure entering the preset start-up pressure range, the turbine inlet temperature reaching the start-up temperature range, the outlet pressure of the pressure buffer container 20 stabilizing, and the turbine inlet regulating valve 29 being within the controllable opening range. The preset start-up pressure range is determined based on the outlet pressure of the low-pressure CO2 storage tank 6, the allowable inlet pressure of the CO2 compression pump group 7, and the pressure required for the high-pressure jet mixer 21 to establish a working pressure difference. In this embodiment, when the main circulation pressure enters the range of 8MPa to 10MPa, the controller determines that the main circulation pressure meets the start-up pressure condition. The start-up temperature range is determined based on the allowable inlet temperature of the turbine 1, the preheating capacity of the regenerative heat exchanger 3, and the outlet temperature of the high-pressure CO2 heater 8. In this embodiment, when the turbine inlet temperature reaches 90℃ to 115℃, the controller determines that the turbine inlet temperature meets the start-up temperature condition. Stable outlet pressure of the pressure buffer container 20 means that the pressure fluctuation does not exceed 0.2MPa within 3 to 5 consecutive control cycles. The controllable opening range of the turbine inlet regulating valve 29 can be taken as 10% to 30% to avoid the turbine inlet regulating valve 29 being in a near-fully closed position or a large-amplitude sudden opening position.

[0051] The controller generates a heat source power prediction result based on power generation control information and historical operating sequences. The feature vector used for prediction is denoted as... ,in Indicates the current sampling time. This includes the time-series values ​​of heat source power over a past period, the current hour, weekday information, weekday markers, ambient temperature, humidity, wind speed, future weather forecasts, upstream production line load rates, planned shutdown signals, maintenance signals, and heat source change rates. The controller... After normalization, the data is input into a Gaussian process regression model to obtain the predicted mean and variance of heat source power in the future prediction time domain.

[0052] like Figure 3As shown, variable operating condition control is achieved through a pressure compensation branch and pressure-flow mapping. The high-pressure supercritical CO2 storage tank 9 supplies high-pressure working fluid to the pressure buffer vessel 20. The pressure buffer vessel 20 supplies working fluid to the high-pressure jet mixer 21 via the injection valve 36. The outlet of the cryogenic working fluid channel of the regenerating heat exchanger 3 supplies the ejected main working fluid to the high-pressure jet mixer 21 via the main check valve 34. The controller does not fix the turbine inlet pressure to a single rated value, but instead forms a working fluid control target based on the available heat source power and ambient temperature. The correspondence between available heat source power and turbine inlet pressure is as follows:

[0053] ;

[0054] in, To increase turbine inlet pressure, The power of the currently available heat source. This is a pressure-flow mapping table or equivalent calculation model. When the ambient temperature changes and cooling conditions change, the controller selects a pressure-flow mapping node from the corresponding family of ambient temperature curves and outputs the turbine inlet pressure target and the CO2 mass flow rate target. Taking an ambient temperature of 25℃ as an example, the controller can use the discrete nodes listed in Table 1 below for piecewise linear interpolation.

[0055] Table 1 Discrete Node Table

[0056]

[0057] The pressure-flow mapping table is established offline using a closed-loop thermodynamic model. The closed-loop thermodynamic model includes a turbine 1, a CO2 compressor pump unit 7, a regenerative heat exchanger 3, a high-pressure CO2 heater 8, a cooling unit 4, and a CO2 property calculation module. The CO2 property calculation module calculates the enthalpy, density, and compressibility factor of the working fluid based on the turbine inlet temperature, turbine inlet pressure, turbine outlet pressure, and the CO2 temperature at the outlet of cooling unit 4. This can be achieved using REFPROP software or a property equation of state adapted to the supercritical CO2 working fluid state range. With the goal of maximizing cycle efficiency, the cycle efficiency can be calculated according to the following relationship:

[0058] ;

[0059] ;

[0060] in, For closed-loop efficiency. CO2 mass flow rate, For turbine output power, The CO2 compression pump unit consumes 7 units of power. This represents the heat absorbed by the high-pressure CO2 heater 8 from the heat source. When solving offline, it is... and As decision variables, the upper limit of turbine inlet temperature safety, surge boundary of CO2 compressor pump unit 7, blockage boundary, and minimum end difference of heat exchanger are used as equipment constraints for different... Find the pressure-flow combination.

[0061] Turbine output power can be calculated as follows:

[0062] ;

[0063] in, The inlet enthalpy of turbine 1. The enthalpy at the turbine outlet is 1. The isentropic efficiency of turbine 1. and Determined by the turbine inlet temperature, turbine inlet pressure, turbine outlet pressure, and CO2 property calculation module. The parameters are determined based on the performance curves, bench test data, or simulation calibration data provided by the turbine manufacturer, and obtained by referring to tables according to the current operating conditions of turbine 1, including inlet pressure, inlet temperature, outlet pressure, and mass flow rate.

[0064] During operation, the controller calculates the measured value of the cycle efficiency. And compared with the predicted cycle efficiency values ​​given by the pressure-flow mapping table. The mapping deviations are as follows:

[0065] ;

[0066] in, This represents the deviation in the cycle efficiency mapping. If... If the deviation exceeds the preset threshold continuously, the controller will use the most recent The group runs the data and corrects the neighborhood mapping nodes of the working conditions using recursive least squares or extreme gradient boosting algorithms. This indicates the number of running data sets used for mapping node correction, determined by the controller's storage window length.

[0067] The Gaussian process regression model uses a composite kernel function, which is as follows:

[0068] ;

[0069] ;

[0070] ;

[0071] ;

[0072] in, and Given two sets of input feature vectors, and These represent the input sample sequence numbers; It is a squared exponent kernel used to describe the smooth relationship between heat source power and operating parameters; The variance of the squared exponential kernel signal; Indicates the input feature dimension index; Indicates the total dimension of the input features; and They are respectively and In the Values ​​can be taken in each input feature dimension; For the first The length scale of each input feature dimension; It serves as a periodic kernel, used to describe daily or weekly process fluctuations; and They are respectively and The periodic time characteristics in; The variance of the periodic kernel signal; The period length; The periodic kernel length scale; It is a white noise kernel; To measure the noise variance; For sample sequence number indication, when When, take 1, when The time parameter is set to 0. These hyperparameters are determined by maximizing the logarithmic marginal likelihood of historical running data.

[0073] The output of the Gaussian process regression model is as follows:

[0074] ;

[0075] in, This indicates a prediction of future moments within the time domain. For the future The heat source power random variable, The predicted average power of the heat source. The variance of heat source power prediction. The distribution represents a normal distribution. The controller collects new data and updates the posterior distribution in each control cycle. When the prediction error exceeds the prediction error threshold or the prediction interval coverage is lower than the preset coverage lower limit, the controller updates the Gaussian process regression model to reconstruct the heat source power prediction result. The prediction error is calculated by the deviation between the actual heat source power and the predicted mean. This can be the absolute deviation between the actual heat source power and the predicted mean, or the ratio of this absolute deviation to the actual heat source power. The prediction error threshold is determined based on the accuracy of the heat source power sensor, the standard deviation of historical prediction residuals, and the control safety margin. In one implementation, the prediction error threshold is three times the standard deviation of historical prediction residuals, or 5% of the rated heat source power. The prediction interval coverage is calculated by the proportion of actual heat source power falling within the prediction confidence interval. The preset coverage lower limit can be between 0.90 and 0.95; in this embodiment, it is 0.95. When the prediction interval coverage is lower than the preset coverage lower limit, it indicates that the current model underestimates the uncertainty of heat source fluctuations, and the controller updates the hyperparameters or training sample window of the Gaussian process regression model.

[0076] The change in heat source power is determined by the difference between the current heat source power and the operating baseline. The heat source regulation targets include a heat storage distribution target and a flash steam replenishment target. The heat storage distribution target is used to adjust the inflow ratio from the hot water inlet valve group 15 to the first heat storage zone 12 and the second heat storage zone 13. The flash steam replenishment target is used to adjust the water intake from the shell-side saturated water space of the first heat storage zone 12 by the flash valve group and to replenish steam to the steam-water mixing heater by the flash steam replenishment device 14. If the change in heat source power is positive, the controller accumulates the positive change within the sliding window to characterize the amplitude of the sustained thermal shock. The accumulated positive change within the sliding window is determined by the following formula:

[0077] ;

[0078] in, For the first The cumulative positive change within the sliding window of each control cycle. The number of control cycles contained in the sliding window. This is the control cycle index within the sliding window. For the first The heat source power is collected in each control cycle. The control cycle length is determined based on the response time of the hot water inlet valve group 15, the thermal inertia of the heat storage tank 10, and the control cycle duration, and can be 3 to 10 control cycles. When the control cycle is 2 seconds, the sliding window corresponds to a heat source change observation time of 6 to 20 seconds. The positive threshold and the downward threshold are determined based on the allowable fluctuation range of the outlet temperature of the heat storage tank 10, the response time of the hot water inlet valve group 15, and the statistical results of the heat source power change rate during the historical stable operation phase. In one implementation, the positive threshold and the downward threshold are respectively taken as the 95th percentile of the absolute value of the heat source power change rate during the historical stable operation phase, plus a 10% safety margin. The cumulative threshold is determined based on the remaining absorbable phase change heat storage of the first heat storage zone 12, the sliding window length, and the control cycle duration, and is used to determine whether the continuous thermal shock in a short period of time has reached the level that requires priority introduction into the first heat storage zone 12.

[0079] If the ratio of the change in heat source power to the duration of the control cycle exceeds the positive threshold, or if the cumulative positive change within the sliding window exceeds the cumulative threshold, the controller generates a heat source regulation target that increases the inflow ratio of the first heat storage zone 12.

[0080] If the change in heat source power is negative, the controller obtains the temperature drop rate of the first heat storage zone 12, which is determined by the following formula:

[0081] ;

[0082] in, For the first The rate of temperature decrease in the first heat storage zone 12 of each control cycle For the first heat storage zone 12 in the Temperature per control cycle, To control the cycle duration, the controller adjusts the flash steam replenishment trigger temperature threshold based on the temperature drop rate of the first heat storage zone 12. The temperature drop rate threshold of the first heat storage zone 12 is determined based on the allowable drop in the target heating temperature and the time required for the flash steam replenishment device 14 to form a stable steam flow rate after receiving the start-up command. Specifically, the allowable drop in the target heating temperature is determined by the hot water inlet temperature requirement of the high-pressure CO2 heater 8 and the safety requirements for the downstream turbine inlet temperature. The response time of the flash steam replenishment device 14 is determined by the flash valve group's operating time, the pressure before and after the valve, and the gas-liquid separation space volume. When the temperature drop rate of the first heat storage zone 12 does not exceed this temperature drop rate threshold, the flash steam replenishment trigger temperature threshold is the target heating temperature. When the temperature drop rate of the first heat storage zone 12 exceeds this temperature drop rate threshold, a feedforward temperature margin is added to the target heating temperature to obtain an improved flash steam replenishment trigger temperature threshold. The feedforward temperature margin can be between 2°C and 5°C; in this embodiment, it is set to 3°C. This margin is used to ensure that the flash steam replenishment action occurs before the outlet water temperature of the steam-water mixing heater falls below the target heating temperature. If the ratio of the absolute value of the heat source power change to the control cycle duration exceeds the descent threshold, and the outlet water temperature of the steam-water mixing heater is lower than the flash steam replenishment trigger temperature threshold, the controller generates a flash steam replenishment target. The controller uses the heat storage distribution target and the flash steam replenishment target as inputs to solve for the control quantity, determining the control quantity for the hot water inlet valve group 15 or the flash valve group, so that the heat source buffering action matches the direction of heat source change.

[0083] The controller establishes safety constraints based on the circulating working fluid state and equipment limits. The controller inputs the predicted mean of heat source power into the thermodynamic model for forward calculation, obtaining the predicted mean of turbine inlet temperature and pressure of high-pressure supercritical CO2 storage tank 9. The controller then transfers the predicted variance of heat source power to the turbine inlet temperature and pressure of high-pressure supercritical CO2 storage tank 9 according to the sensitivity or unscented transformation of the thermodynamic model at the current operating point, obtaining the predicted standard deviations of these two pressures. In the first-order Taylor expansion, the sensitivity of the state variables to heat source power is determined by the partial derivatives of the thermodynamic model near the current operating point, and the predicted variance is converted into the state variable variance after being weighted by the square of the sensitivity.

[0084] like Figure 4As shown, the cold-end cold storage structure is used to distribute power generation revenue between power generation revenue and cold storage revenue. The hot water inlet of the hot water lithium bromide chiller unit 5 and the hot water inlet of the high-pressure CO2 heater 8 are both connected to hot water supply pipelines led from a steam-water mixing heater. The chilled water outlet of the hot water lithium bromide chiller unit 5 is connected to the chilled water storage tank 25. The outlet of the chilled water storage tank 25 is connected to the inlet of the cold storage distribution three-way valve 26. The first outlet of the cold storage distribution three-way valve 26 is connected to the subcooling heat exchanger 27, which is connected in series between the outlet of the cooling unit 4 and the inlet of the low-pressure CO2 storage tank 6. The second outlet of the cold storage distribution three-way valve 26 is connected to the auxiliary cooling branch, which is connected to the turbine oil cooler and the generator cooling water jacket. When chilled water enters the subcooling heat exchanger 27, it lowers the CO2 temperature before entering the low-pressure CO2 storage tank 6. When chilled water enters the auxiliary cooling branch, it cools the turbine oil cooler and the generator cooling water jacket. The cold storage distribution three-way valve 26 is connected to the controller signal. The controller adjusts the cold storage distribution three-way valve 26 to adjust the cold water distribution ratio from the chilled water storage tank 25 to the subcooling heat exchanger 27 and the auxiliary cooling branch.

[0085] The controller calculates both instantaneous power generation revenue and delayed cold storage revenue within the same forecast time domain. The revenue metrics are as follows:

[0086] ;

[0087] ;

[0088] in, For real-time power generation revenue indicators, For power generation capacity, At the current electricity price, For the benefit of delayed cold storage, The value representing the efficiency improvement of a closed-loop system after using chilled water for deep subcooling of the working fluid. This represents the maximum power output of the power generation system. For future peak-hour electricity prices, This refers to the duration of future peak hours. Turbine output power It was obtained by converting the generator efficiency.

[0089] To ensure that the revenue comparisons are conducted at the same time, the controller can also calculate the cumulative revenue over the prediction time domain as follows:

[0090] ;

[0091] ;

[0092] in, To predict the cumulative revenue from real-time power generation within the time domain, The electricity price for the current period is... To control the cumulative duration of the cycle within the prediction time domain, To accumulate revenue from extended cold storage, This is the cost item after converting the operating power consumption cost, start-up and shutdown loss cost, and operating correction amount of the refrigeration unit. and All calculations are performed using the same currency unit and the same forecast time domain.

[0093] The controller determines the cold storage threshold based on the cumulative revenue from real-time power generation and a preset revenue margin coefficient. The preset revenue margin coefficient is denoted as... , The value is determined by electricity price fluctuations, chiller start-up and shutdown costs, and operating strategies, and ranges from [value range missing]. The cold storage threshold can be calculated according to... Confirmed; Cumulative benefits of delayed cold storage When the cold storage threshold is reached, the controller allows the hot water-type lithium bromide chiller unit 5 to store cold water in the chilled water storage tank 25. When power generation priority is determined based on the power generation control information, the controller limits the hot water flow of the hot water-type lithium bromide chiller unit 5, allowing hot water to preferentially enter the high-pressure CO2 heater 8. If power generation priority is currently prioritized, the CO2 temperature at the outlet of the cooling unit 4 exceeds the target value, and the available cold capacity in the chilled water storage tank 25 is higher than the lower limit for cold release, the controller prioritizes adjusting the cold storage distribution three-way valve 26, allowing the stored chilled water in the chilled water storage tank 25 to supply cooling to the subcooling heat exchanger 27, and can also supply cooling to the turbine oil cooler and generator cooling water jacket through the auxiliary cooling branch. The available cold capacity in the chilled water storage tank 25 is determined based on the mass of chilled water available for cold release, the specific heat capacity of the water, and the temperature difference between the chilled water supply and return. The lower limit for cold release is determined based on the rated available cold capacity of the chilled water storage tank 25 and the cold-end safety margin, and can be taken as 10% to 20% of the rated available cold capacity; in this embodiment, it is taken as 15%. When the available cooling capacity in the chilled water storage tank 25 is lower than the cooling capacity limit, the controller will no longer use the chilled water in the chilled water storage tank 25 for the additional subcooling demand of the subcooling heat exchanger 27, but will only reserve it for necessary auxiliary cooling of the turbine oil cooler and generator cooling water jacket. During this process, the controller will not start the hot water lithium bromide chiller unit 5 or increase the hot water flow of the hot water lithium bromide chiller unit 5 due to the cooling demand of the subcooling heat exchanger 27 or the auxiliary cooling branch, and will only meet the subcooling and auxiliary cooling needs without consuming additional hot water.

[0094] like Figure 5As shown, the redundant crossover protection structure includes three functionally decoupled crossover pipelines. The first is the compressor return pipeline, connected between the outlet main of the CO2 compressor pump group 7 and the gas phase zone of the low-pressure CO2 storage tank 6, and equipped with a first fast regulating valve 30; the second is the cold flow bypass, connected between the working fluid inlet and outlet pipeline of the high-pressure CO2 heater 8, and equipped with a second fast regulating valve 31, where the low-temperature working fluid drawn from the cold flow bypass mixes with the heated high-temperature working fluid at the outlet side of the high-pressure CO2 heater 8; the third is the emergency pressure reduction bypass, connected between the pipeline section before the turbine inlet regulating valve 29 and the pipeline section between the turbine 1 outlet and the high-temperature working fluid channel inlet of the regenerating heat exchanger 3, and connected in series with a quick-opening solenoid valve 32 and a multi-stage throttling orifice plate 33. The first fast regulating valve 30 and the second fast regulating valve 31 can be fast-response electric regulating valves, and the quick-opening solenoid valve 32 can be a pneumatic quick-opening solenoid valve. The first fast regulating valve 30, the second fast regulating valve 31, and the quick-opening solenoid valve 32 all receive the linkage control command output by the controller.

[0095] When the forward main flow rate of the closed loop is... When the flow rate drops below the minimum continuous stable flow rate, the controller opens the first quick regulating valve 30, allowing a portion of the working fluid to flow at a return flow rate. The gas phase of the CO2 is returned to the low-pressure CO2 storage tank 6 to prevent the CO2 compression pump unit 7 from entering the low-load surge zone. The minimum continuous stable flow rate is determined by the performance curve of CO2 compressor pump set 7, collected by the outlet flow meter. Specifically, the controller reads the surge boundary flow rate of CO2 compressor pump set 7 at the current speed and inlet condition, and adds a flow safety margin of 5% to 10% to the surge boundary flow rate to form the minimum continuous stable flow rate. When the complete performance curve is not obtained, the minimum continuous stable flow rate can be taken as 15% to 30% of the rated flow rate of CO2 compressor pump set 7. After the current inbound main flow rate meets the safe operation requirements, the controller reduces the opening of the first fast regulating valve 30; when the current inbound main flow rate is lower than the minimum continuous stable flow rate again, the controller increases the opening of the first fast regulating valve 30 again, so that part of the working fluid flows back to the gas phase zone of the low-pressure CO2 storage tank 6. When the turbine inlet temperature exceeds the safety upper limit, the controller opens the second fast regulating valve 31, so that part of the low-temperature working fluid on the inlet side of the high-pressure CO2 heater 8 bypasses the high-pressure CO2 heater 8 and mixes with the heated working fluid at the mixing point on the outlet side. When turbine 1 receives an emergency stop signal, the controller opens the quick-opening solenoid valve 32. The high-pressure CO2 is depressurized by the multi-stage throttling orifice plate 33 and then enters the inlet pipe section of the high-temperature working medium channel of the regenerating heat exchanger 3 from the outlet of turbine 1.

[0096] The controller sets chance constraints on the pressure of the high-pressure supercritical CO2 storage tank 9 and the turbine inlet temperature in the prediction time domain. The chance constraints are as follows:

[0097] ;

[0098] ;

[0099] in, Represents probability. Indicates the step number in the prediction time domain. Indicates the number of prediction time-domain steps. For the first Predicted pressure of high-pressure supercritical CO2 storage tank 9 For the first Step allowable pressure limit, For the first Predicted turbine inlet temperature For the first Step allowable temperature limit, Pre-set reliability. This is used to limit the probability requirement that the predicted pressure and temperature are within the safe upper limit, and the value can range from 0.95 to 0.99. When the predicted variance of the heat source power increases, the turbine inlet temperature approaches the safe upper limit, or the pressure of the high-pressure supercritical CO2 storage tank 9 approaches the safe upper limit, the controller will... The value is set to 0.98 to 0.99; when the heat source power fluctuation decreases and the system is operating at a non-full load, the controller will... The value is taken to be between 0.95 and 0.97. In this embodiment, Take 0.95.

[0100] When the state prediction error is treated according to a Gaussian distribution, the turbine inlet temperature constraint can be transformed into a deterministic boundary constraint:

[0101] ;

[0102] ;

[0103] ;

[0104] in, The cumulative distribution function of the standard normal distribution. For a standard normal distribution at a confidence level quantiles at , For the first Predicted mean inlet temperature of stepper turbine. For the first Standard deviation of turbine inlet temperature prediction To ensure the safe upper limit of turbine inlet temperature, For the first Average pressure prediction for step 9 high-pressure supercritical CO2 storage tanks For the first Standard deviation of pressure prediction for step-high pressure supercritical CO2 storage tank 9 The pressure safety limit for high-pressure supercritical CO2 storage tank 9.

[0105] like Figure 6 As shown, the controller employs a hierarchical control structure comprising stochastic heat source prediction, chance-constrained rolling optimization, and real-time feedback correction. The stochastic heat source prediction section generates a probability distribution of heat source power based on historical heat source data, meteorological conditions, and process characteristics. The chance-constrained rolling optimization section incorporates revenue objectives, equipment constraints, thermal storage status, heat release rate, CO2 flow rate, pressure, and chilled water consumption into the same optimization problem, transforming chance constraints into deterministic boundary constraints. The real-time feedback correction section sends the first control variable to the valves, pumps, and CO2 compressor pump group 7, while simultaneously collecting actual heat source power, temperature, pressure, and flow rate data to update the Gaussian process regression model and state estimation model. The rolling optimization aims to maximize power generation revenue, match thermal storage revenue, minimize temperature and pressure exceedance penalties, and control the actuator's movement amplitude. The controller recalculates the control variable sequence within the prediction time domain at each control cycle.

[0106] The controller combines the predicted heat source power, heat source regulation target, and working fluid control target to solve for the control quantities. These control quantities include the CO2 compressor pump unit 7 speed, turbine inlet regulating valve 29 opening degree, pressure regulating valve 35 opening degree, jet valve 36 opening degree, hot water inlet valve group 15 opening degree, flash evaporation valve group opening degree, cold storage distribution three-way valve 26 opening degree, first fast regulating valve 30 opening degree, second fast regulating valve 31 opening degree, and the open / closed state of quick-opening solenoid valve 32. Within each control cycle, the controller solves for the control quantity sequence within the prediction time domain, outputting only the first control quantity in the sequence; in the next control cycle, data is re-acquired, prediction results are updated, and the control quantities are solved again. The prediction time domain can include nodes such as 15min, 30min, 45min, and 60min, and the control cycle can be set according to the actuator's response capability.

[0107] In the power generation priority control, if the controller determines that the current period is a peak electricity price period or the cumulative revenue from immediate power generation is higher than the cumulative revenue from delayed cold storage, it limits the hot water flow of the hot water-type lithium bromide chiller unit 5, so that the hot water preferentially enters the high-pressure CO2 heater 8. When the CO2 temperature at the outlet of the cooling unit 4 exceeds the target value and the available cooling capacity in the chilled water storage tank 25 is higher than the lower limit of cooling release, the controller prioritizes adjusting the cold storage distribution three-way valve 26, so that the chilled water storage tank 25 supplies cooling to the subcooling heat exchanger 27, and avoids additional hot water consumption for this subcooling demand. The target value of the CO2 temperature at the outlet of the cooling unit 4 is determined according to the inlet temperature requirements of the low-pressure CO2 storage tank 6, the working fluid state at the inlet of the CO2 compression pump group 7, and the design operating conditions of the cooling unit 4. In this embodiment, when the low-pressure CO2 storage tank 6 is operating at 8MPa and 20℃, the target value of the CO2 temperature at the outlet of the cooling unit 4 can be set to 20℃ to 25℃. When the ambient temperature rises and causes the CO2 temperature at the outlet of cooling unit 4 to exceed the target value, the controller prioritizes using the stored cooling capacity in the chilled water storage tank 25 for subcooling, without increasing the hot water flow rate of the hot water-type lithium bromide chiller unit 5 due to this subcooling demand. If power generation is not the priority and the cumulative benefit of delayed cold storage reaches the cold storage threshold, the controller allows the hot water-type lithium bromide chiller unit 5 to store cold in the chilled water storage tank 25.

[0108] In redundancy protection control, if the controller detects... If the flow rate falls below the minimum continuous stable flow rate, the first quick-acting regulating valve 30 is opened to allow the compressor pump return pipe to open. If the turbine inlet temperature exceeds the safety upper limit, the second quick-acting regulating valve 31 is opened to allow the cold flow bypass to open. If the turbine 1 receives an emergency stop signal, the quick-opening solenoid valve 32 is opened to allow the emergency pressure reduction bypass to open. The quick-opening solenoid valve 32 enters a latching protection state after the emergency pressure reduction bypass is opened. After the emergency stop signal is released, the controller first determines whether the turbine inlet pressure, the pressure of the high-pressure supercritical CO2 storage tank 9, the turbine inlet temperature, and the main flow rate have returned to the reset allowable range. The reset allowable range includes: turbine inlet pressure below the safe upper limit of turbine inlet pressure with a pressure hysteresis of 0.5MPa to 1.0MPa; pressure in high-pressure supercritical CO2 storage tank 9 below the safe upper limit of high-pressure supercritical CO2 storage tank 9 with a pressure hysteresis of 1.0MPa to 2.0MPa; turbine inlet temperature below the safe upper limit of turbine inlet temperature with a temperature hysteresis of 5°C to 10°C; and main pipeline mass flow rate restored to above the minimum continuous stable flow rate. After the reset allowable range is met, the controller continues to wait for a preset reset delay, which can be 5s to 30s, and is 10s in this embodiment. When all parameters remain within the reset allowable range within the preset reset delay, the controller closes the quick-opening solenoid valve 32 and releases the emergency pressure reduction bypass conduction according to the preset reset method. The preset reset methods include manual confirmation reset and automatic reset after the interlock conditions are met. Manual confirmation reset is used for emergency pressure reduction triggered by the emergency stop signal of turbine 1, protection interlock fault, or operator stop command. Automatic reset is used for emergency pressure reduction after the heat source transient disturbance is resolved and no fault latch-up has been formed. The three protection branches correspond to low-load surge, heat source overheating, and load shedding overpressure conditions, respectively, and are uniformly incorporated into the linkage control commands by the controller. In the next control cycle after the protection branch is turned on or reset, the controller re-collects the heat source power, turbine inlet temperature, main line mass flow rate, and pressure of the high-pressure supercritical CO2 storage tank 9, and re-forms the heat source power prediction results and control quantities.

[0109] In this embodiment, the parameters include: forward threshold, descent threshold, sliding window length, cumulative threshold, first heat storage zone temperature descent rate threshold, feedforward temperature margin, preset coverage lower limit, and preset confidence level. The preset revenue margin coefficient λ, chilled water storage tank 25 cooling lower limit, minimum continuous stable flow rate, target heating temperature, turbine inlet temperature safety upper limit, high-pressure supercritical CO2 storage tank 9 pressure safety upper limit, emergency pressure reduction bypass reset allowable range, preset reset delay, and preset reset method can all be stored in the controller parameter table and determined based on the equipment's factory performance curve, on-site commissioning data, operating economic strategies, and safety regulations. The controller continuously regulates the supercritical CO2 power generation process under drastically fluctuating industrial waste heat conditions through valve opening, pump speed, jet pressure, heat source distribution, cold storage distribution, and bypass protection actions.

[0110] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat, comprising a closed-loop circuit, a heat source buffer branch, a pressure compensation branch, and a controller, characterized in that, In the closed-loop circuit, the outlet of the low-pressure CO2 storage tank is connected to the injection inlet of the high-pressure jet mixer via the CO2 compression pump set, the low-temperature working fluid channel of the regenerating heat exchanger, and the check valve. The outlet of the high-pressure jet mixer is connected back to the low-pressure CO2 storage tank via the high-pressure CO2 heater, the turbine inlet regulating valve, the turbine, the high-temperature working fluid channel of the regenerating heat exchanger, and the cooling unit. In the heat source buffer branch, the hot water inlet valve group is connected to the heat storage tank. The hot water outlet of the heat storage tank is connected to the hot water inlet of the high-pressure CO2 heater via the steam-water mixing heater. The flash steam replenishment device is equipped with a flash valve group and its steam outlet is connected to the steam inlet of the steam-water mixing heater. In the pressure compensation branch, the high-pressure supercritical CO2 storage tank with an operating pressure higher than the target pressure at the outlet of the high-pressure jet mixer is connected to the working fluid inlet of the high-pressure jet mixer via the pressure regulating valve, the pressure buffer container, and the jet valve. The controller is connected to the CO2 compression pump set, the turbine inlet regulating valve, the pressure regulating valve, the jet valve, the hot water inlet valve group, and the flash valve group via signal connections. The heat storage tank is divided into a first heat storage zone and a second heat storage zone. The hot water inlet valve group is connected to the first heat storage zone and the second heat storage zone. The first heat storage zone is equipped with a sealed phase change heat storage unit and a shell-side saturated water space. The flash steam replenishment device draws water from the shell-side saturated water space through the flash steam valve group. The second heat storage zone is equipped with a hot water storage space and a heat replenishment interface. The system also includes a compressor return pipe with a first fast-acting regulating valve, a cold flow bypass with a second fast-acting regulating valve, and an emergency pressure reduction bypass with a quick-opening solenoid valve and a multi-stage throttling orifice plate. The compressor return pipe is connected between the outlet of the CO2 compressor pump group and the gas phase zone of the low-pressure CO2 storage tank. The cold flow bypass is connected between the inlet and outlet of the high-pressure CO2 heater working fluid. The emergency pressure reduction bypass is connected between the pipe section before the turbine inlet regulating valve and the turbine outlet pipe section. The first fast-acting regulating valve, the second fast-acting regulating valve, and the quick-opening solenoid valve are all connected to the controller signal.

2. The supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat according to claim 1, characterized in that, It also includes a hot water lithium bromide chiller unit, a chilled water storage tank, a cold storage distribution three-way valve, a subcooling heat exchanger, and auxiliary cooling branches; the hot water inlet of the hot water lithium bromide chiller unit and the hot water inlet of the high-pressure CO2 heater are both connected to the hot water supply pipeline led from the steam-water mixing heater, and the chilled water outlet of the hot water lithium bromide chiller unit is connected to the chilled water storage tank; the chilled water storage tank is connected to the subcooling heat exchanger and the auxiliary cooling branches respectively via the cold storage distribution three-way valve, the subcooling heat exchanger is connected in series between the cooling unit outlet and the low-pressure CO2 storage tank inlet, and the auxiliary cooling branches are connected to the turbine oil cooler and the generator cooling water jacket; the cold storage distribution three-way valve is connected to the controller signal.

3. A control method applied to a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat, as described in any one of claims 1 to 2, characterized in that, include: S101: Based on the operating status of the closed-loop circuit, heat source buffer branch, and pressure compensation branch, combined with power generation control information and historical operating sequences, identify the operating mode and form an operating benchmark; S102: Based on the power generation control information and historical operating sequence, the heat source power prediction result is formed. The difference between the current heat source power and the operating benchmark is used to form the heat source power change, and heat source regulation target is generated for regulating the heat storage tank and flash steam injection device. S103: Based on the available heat source power and ambient temperature, a working fluid control target is formed, which includes a turbine inlet pressure target and a CO2 mass flow rate target. S104: Safety constraints are formed by the state of the circulating working fluid and equipment limits. The control quantity is solved by combining the heat source power prediction results, heat source adjustment target and working fluid control target, and the linkage control command is output to the CO2 compression pump group, turbine inlet regulating valve, pressure regulating valve, jet valve, hot water inlet valve group and flash evaporation valve group.

4. The control method for a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat according to claim 3, characterized in that, In step S101, when the start command and heat source availability signal are valid, the CO2 compression pump group is in an unstarted state, and the main circulation pressure has not reached the start pressure, the operating mode is determined to be the black start mode; in step S104, in the black start mode, the pressure regulating valve and the injection valve are opened, so that the high-pressure supercritical CO2 storage tank supplies pressure to the working fluid inlet of the high-pressure CO2 heater through the pressure buffer container and the high-pressure injection mixer; after the state of the circulating working fluid meets the start conditions, the CO2 compression pump group is started to establish the main circulation, so that the high-pressure injection mixer enters the ejector mixing state.

5. The control method for a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat according to claim 3, characterized in that, In step S102, the historical data of the heat source, meteorological and process characteristics are input into a Gaussian process regression model containing a squared exponential kernel, a periodic kernel and a white noise kernel to obtain the predicted mean and variance of the heat source power. When the prediction error exceeds the prediction error threshold or the prediction interval coverage is lower than the preset coverage lower limit, the Gaussian process regression model is updated to regenerate the heat source power prediction results.

6. The control method for a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat according to claim 3, characterized in that, In step S102, the heat source regulation target includes a heat storage distribution target and a flash steam replenishment target. The heat storage distribution target is used to adjust the inflow ratio of the hot water inlet valve group to the first heat storage zone and the second heat storage zone in the heat storage tank. The flash steam replenishment target is used to adjust the flash valve group to draw water from the shell-side saturated water space of the first heat storage zone and to replenish steam to the steam-water mixing heater by the flash steam replenishment device. When the change in heat source power is positive and the ratio of the change in heat source power to the control cycle duration exceeds the positive threshold, or the cumulative positive change in the sliding window exceeds the cumulative threshold, a heat storage distribution target that increases the inflow ratio of the first heat storage zone is generated. When the change in heat source power is negative and the ratio of its absolute value to the control cycle duration exceeds the descent threshold, and the outlet water temperature of the steam-water mixing heater is lower than the flash steam replenishment trigger temperature threshold after the temperature descent rate correction of the first heat storage zone, a flash steam replenishment target is generated; in step S104, the heat storage distribution target and the flash steam replenishment target are used as inputs to solve the control quantity, and the control quantity of the hot water inlet valve group or the flash valve group is determined.

7. The control method for a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat according to claim 3, characterized in that, In step S103, based on a closed-loop thermodynamic model including a CO2 compression pump group, a regenerative heat exchanger, a high-pressure jet mixer, a high-pressure CO2 heater, a turbine, and a cooling unit, with the goal of maximizing cycle efficiency, the correspondence between available heat source power, ambient temperature, turbine inlet pressure, and CO2 mass flow rate is solved offline under equipment constraints to construct a pressure-flow mapping table. Based on the deviation between the measured and predicted values ​​of cycle efficiency, the recursive least squares method or extreme gradient boosting algorithm is used to correct the neighborhood mapping nodes of the operating conditions.

8. The control method for a supercritical CO2 power generation system adapted to drastically fluctuating industrial waste heat according to claim 5, characterized in that, In step S104, the predicted mean of heat source power is input into a thermodynamic model containing a heat storage tank, a steam-water mixing heater, a high-pressure CO2 heater, a high-pressure supercritical CO2 storage tank, a pressure buffer vessel, a high-pressure jet mixer, and a turbine for forward calculation. The predicted variance of heat source power is then transferred to the turbine inlet temperature and the high-pressure supercritical CO2 storage tank pressure according to the sensitivity or unscented transformation of the thermodynamic model at the current operating point, to obtain the predicted mean and standard deviation of the turbine inlet temperature and the high-pressure supercritical CO2 storage tank pressure. The predicted mean and standard deviation are then converted into deterministic boundary constraints according to a preset confidence level. In each control cycle, the control quantity sequence in the prediction time domain is solved, and the first control quantity in the control quantity sequence is output.

Images