Power plant circulating water waste heat utilization self-adaptive regulation and control method based on dynamic heat source characteristics

Abstract

The invention relates to a power plant circulating water waste heat utilization self-adaptive regulation and control method based on dynamic heat source characteristics, which comprises the following steps that distributed detection units are arranged on circulating water supply and return water pipelines, the top of a condenser and the secondary side of a heat supply network, and the circulating water supply and return water temperature and flow, the condenser vacuum degree and the heat supply network supply and return water temperature are collected in real time; on the main machine side, the opening degree change step length of a condenser adjusting valve is smaller than or equal to 5% / s, and the signal delay is smaller than or equal to 100 ms, and on the heat pump side, a load-circulating water parameter correlation prediction model is constructed, in combination with the requirements of main machine safety, heat pump efficiency and terminal requirements with the vacuum degree smaller than or equal to 7 kPa, an adaptive instruction is generated through operation, and the instruction in S3 is converted into a signal capable of being recognized by an executing mechanism; on the heat pump side, the adjustment step length of the solution pump / refrigerant pump is smaller than or equal to 100 r / min, 20 m < 3 > / h And at the terminal side, the heat supply network water mixing valve response time is less than or equal to 2s, and terminal load fluctuation is adapted.

Description

Technical Field

[0001] This invention relates to the field of power plant waste heat utilization technology, and more particularly to an adaptive control method for power plant circulating water waste heat utilization based on dynamic heat source characteristics. Background Technology

[0002] Guided by the "dual carbon" goals, my country's installed capacity of new energy (wind power and photovoltaic) has continued to surge, and in 2024, the proportion of domestic new energy power generation exceeded 25%. The intermittent and fluctuating output of new energy has led to a sharp increase in the pressure on the power grid for peak shaving. As the core peak shaving power source, thermal power units need to operate flexibly with varying loads within the range of 30%-100% of rated load, which directly causes dynamic fluctuations in key parameters of circulating water, such as temperature of 44.5-55℃ and flow rate of ±10%.

[0003] As the largest low-grade waste heat resource in thermal power plants, circulating water waste heat accounts for over 60% of total energy consumption losses, and its efficient recovery is the core path to improving energy utilization. However, existing technologies have significant shortcomings: for example, while CN111706897B achieves cross-seasonal soil heat storage, it uses fixed heat pump operating parameters and fails to adapt to circulating water fluctuations; CN113357692A improves efficiency through dual-cycle heat pumps, but lacks a collaborative protection mechanism between the main unit and the heat pump; WO2015127572A1 focuses on power peak shaving but fails to address the problem of lag in circulating water parameter prediction; CN106989430B's heat network adaptation relies on manual adjustment, resulting in a delayed response; and CN110173735A's self-feedback control is based solely on real-time temperature and fails to achieve advance prediction. These technologies lead to low waste heat utilization efficiency, high main unit safety risks, and an inability to meet the stable operation requirements under dynamic conditions, necessitating an adaptive control method to address these industry pain points. Summary of the Invention

[0004] To address the aforementioned technical problems, an adaptive control method for utilizing waste heat from power plant circulating water based on dynamic heat source characteristics is provided.

[0005] The technical means employed in this invention are as follows: An adaptive control method for utilizing waste heat from power plant circulating water based on dynamic heat source characteristics includes the following steps: S1. Distributed detection units are deployed in the circulating water supply and return pipelines, the top of the condenser, and the secondary side of the heating network to collect circulating water supply and return water temperature and flow rate, condenser vacuum degree, and heating network supply and return water temperature in real time. The collected signals are processed according to the operating conditions: the circulating water flow rate fluctuation signal ±10% is used to remove outliers using the 3σ principle, and the missing signal of heating network temperature ≤30s is filled with interpolation of the same type of load. After processing, the signal is transmitted to the central controller through the industrial bus. S2: Based on the GRU neural network, a load-circulating water parameter correlation prediction model is constructed. The linkage signal of unit load-circulating water temperature-flow rate in S1 is input to predict the fluctuation trend of circulating water temperature and flow rate in the next 15-30 minutes. When the predicted temperature is <46℃ or the flow rate decrease is >8%, a fluctuation warning signal is output to trigger the heat pump pre-adjustment logic in advance. S3: Taking into account the requirements of host safety (vacuum degree ≤7kPa), heat pump efficiency, and end-user needs, an adaptation command is generated through calculation: Heat pump control: When the circulating water temperature is 48-55℃, control the heat pump solution pump speed to 3000-3500 r / min and the refrigerant pump flow rate to 360-400 m³ / min. 3 / h; at a temperature of 44.5-48℃, the output speed is increased to 3500-4500 r / min and the flow rate to 400-500 m³ / h. 3 / h, maintaining heat pump efficiency COP≥1.7; Main unit protection: When the vacuum degree is ≥6.5kPa, reduce the heat pump heat extraction by 5%-10% and simultaneously output a signal to increase the cooling tower fan speed; when the vacuum degree is ≥7kPa, cut off 30% of the heat extraction. Terminal adaptation: When the heating network water supply temperature is <75℃, increase the output signal of the mixing valve by 2-5%; S4: Convert the S3 instruction into a signal that the actuator can recognize. Main unit side: Condenser regulating valve opening change step ≤ 5% / s, signal delay ≤ 100ms; Heat pump side: Solution pump / refrigerant pump adjustment step ≤ 100 r / min, 20 m 3 / h; Terminal side: The response time of the heating network mixing valve is ≤2s, adapting to terminal load fluctuations; S5: Operating condition feedback closed loop, the feedback unit collects the following parameters every 1 second: host vacuum degree confirmation ≤7kPa, heat pump COP confirmation ≥1.7, heating network temperature fluctuation confirmation ≤±1℃, and generates a deviation signal to be sent back to the central controller.

[0006] Furthermore, after step S5, step S6 is also included: iterative optimization of operating conditions. Daily data on variable load control and actual field effects are added to the database. The prediction model is retrained within 24 hours, and the circulating water temperature prediction threshold, heat pump and main unit linkage parameters are updated under different loads, so that the control logic can continuously adapt to the unit's variable load operating conditions.

[0007] Furthermore, in step S1, the deployment of the detection unit is adapted to the operating conditions, with the circulating water pipeline sensor avoiding the turbulent zone and the condenser vacuum sensor installed vertically.

[0008] Furthermore, in step S2, the training logic of the prediction model is adapted to the variable load characteristics: the operating conditions are divided into intervals according to the low load of the unit (30%-50%), the medium load (50%-80%), and the high load (80%-100%), and each interval is trained independently as a sub-model.

[0009] Furthermore, in step S3, the calculation logic of the heat pump regulation is to establish a correlation equation between circulating water temperature, pump speed, and COP to ensure that the COP is stable at 1.7±0.1 at different temperatures.

[0010] Compared with existing technologies, this invention has the following advantages: Through a closed-loop control scheme, the COP stability is reduced from the existing ±15% to ±5%, stabilizing at 1.7±0.1. It boasts a high waste heat recovery rate; a 600MW unit can recover more than 2000 tons of standard coal annually through waste heat recovery. Waste heat utilization efficiency is improved by 30% compared to traditional fixed control. The number of times the condenser vacuum exceeds the standard is reduced from an average of 2-3 times per year to zero, avoiding single-time downtime losses exceeding one million yuan. This invention features full-process automation, minimizing manual intervention time and reducing heating network operation and maintenance costs. It is suitable for heating in northern regions and industrial heating in southern regions, helping the grid absorb wind / solar power and reducing wind and solar curtailment. Detailed Implementation

[0011] It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of this invention can be combined with each other. The technical solutions in the embodiments of this invention are clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0012] This invention discloses an adaptive control method for utilizing waste heat from power plant circulating water based on dynamic heat source characteristics, comprising the following steps: S1. Distributed detection units are deployed in the circulating water supply and return pipelines, the top of the condenser, and the secondary side of the heating network to collect circulating water supply and return water temperature and flow rate, condenser vacuum degree, and heating network supply and return water temperature in real time. The collected signals are processed according to the operating conditions: the circulating water flow rate fluctuation signal ±10% is used to remove outliers using the 3σ principle, and the missing signal of heating network temperature ≤30s is filled with interpolation of the same type of load. After processing, the signal is transmitted to the central controller through the industrial bus. S2: Based on the GRU neural network, a load-circulating water parameter correlation prediction model is constructed. The linkage signal of unit load-circulating water temperature-flow rate in S1 is input to predict the fluctuation trend of circulating water temperature and flow rate in the next 15-30 minutes. When the predicted temperature is <46℃ or the flow rate decrease is >8%, a fluctuation warning signal is output to trigger the heat pump pre-adjustment logic in advance. S3: Taking into account the requirements of host safety (vacuum degree ≤7kPa), heat pump efficiency, and end-user needs, an adaptation command is generated through calculation: Heat pump control: When the circulating water temperature is 48-55℃, control the heat pump solution pump speed to 3000-3500 r / min and the refrigerant pump flow rate to 360-400 m³ / min. 3 / h; at a temperature of 44.5-48℃, the output speed is increased to 3500-4500 r / min and the flow rate to 400-500 m³ / h. 3 / h, maintaining heat pump efficiency COP≥1.7; Main unit protection: When the vacuum degree is ≥6.5kPa, reduce the heat pump heat extraction by 5%-10% and simultaneously output a signal to increase the cooling tower fan speed; when the vacuum degree is ≥7kPa, cut off 30% of the heat extraction. Terminal adaptation: When the heating network water supply temperature is <75℃, increase the output signal of the mixing valve by 2-5%; S4: Convert the S3 instruction into a signal that the actuator can recognize. Main unit side: Condenser regulating valve opening change step ≤ 5% / s, signal delay ≤ 100ms; Heat pump side: Solution pump / refrigerant pump adjustment step ≤ 100 r / min, 20 m 3 / h; Terminal side: The response time of the heating network mixing valve is ≤2s, adapting to terminal load fluctuations; S5: Operating condition feedback closed loop, the feedback unit collects the following parameters every 1 second: host vacuum degree confirmation ≤7kPa, heat pump COP confirmation ≥1.7, heating network temperature fluctuation confirmation ≤±1℃, and generates a deviation signal to be sent back to the central controller.

[0013] Furthermore, after step S5, step S6 is also included: iterative optimization of operating conditions. Daily data on variable load control and actual field effects are added to the database. The prediction model is retrained within 24 hours, and the circulating water temperature prediction threshold, heat pump and main unit linkage parameters are updated under different loads, so that the control logic can continuously adapt to the unit's variable load operating conditions.

[0014] Furthermore, in step S1, the deployment of the detection unit is adapted to the operating conditions, with the circulating water pipeline sensor avoiding the turbulent zone and the condenser vacuum sensor installed vertically.

[0015] Furthermore, in step S2, the training logic of the prediction model is adapted to the variable load characteristics: the operating conditions are divided into intervals according to the low load of the unit (30%-50%), the medium load (50%-80%), and the high load (80%-100%), and each interval is trained independently as a sub-model.

[0016] Furthermore, in step S3, the calculation logic of the heat pump regulation is to establish a correlation equation between circulating water temperature, pump speed, and COP to ensure that the COP is stable at 1.7±0.1 at different temperatures.

[0017] Specifically, the GRU neural network prediction model has a two-layer structure, including a hidden layer with 64 neurons, using tanh as the activation function. The model training data comes from historical operating data of the target power plant for at least one complete heating season, containing over 10,000 time-series samples. Sample features include unit load, ambient temperature, circulating water supply and return temperatures and flow rates, and condenser vacuum. The model uses a sliding window with a time step of 30 minutes to organize the data, uses mean squared error as the loss function, and is trained using the Adam optimizer to predict the circulating water temperature and flow rate trends for the next 15-30 minutes, with a prediction mean absolute error (MAE) of less than 0.5℃.

[0018] The correlation equation for heat pump regulation is obtained based on regression fitting of historical operating data of the heat pump unit, and the correlation equation is: Heat pump performance coefficient COP = k1 × T 0.8 ×v -0.3 (k1 is the fitting constant for the operating conditions, k1=0.0025 for low load, 0.003 for medium load, and 0.0035 for high load), T is the average temperature of the circulating water supply and return, which is the arithmetic mean of the supply and return temperatures collected in S1, and v is the speed of the heat pump solution pump.

[0019] k1 is a fitting constant related to the unit's load conditions, determined through data fitting. For example, k1 is approximately 0.0025 at low load (30%-50%), approximately 0.0030 at medium load (50%-80%), and approximately 0.0035 at high load (80%-100%). In actual control, the central controller uses this equation to solve for the required solution pump speed v based on the predicted circulating water temperature T and the target COP (≥1.7), and then generates the corresponding speed control command. For example, when the predicted circulating water temperature is 45℃ and the unit is under low load conditions, the calculation using the equation requires increasing the pump speed to approximately 3600 r / min to ensure that the COP remains stable above 1.7.

[0020] In the implementation of this invention, the power plant side serves as the core link for waste heat generation and preliminary extraction. The process revolves around turbine exhaust steam cooling, circulating water heat exchange, and waste heat diversion and extraction, fully adapting to the unit's 30%-100% variable load operation requirements. In actual production, the 45-55℃ exhaust steam discharged from the turbine is transported to the condenser via pipeline, where it exchanges heat with 20-25℃ circulating water supplied by the circulating water pump. The exhaust steam condenses into water for recycling, while the circulating water absorbs heat and rises to 28-36℃. At this point, the circulating water is divided into two paths: the main path is transported along the pipeline to the cooling tower, where it is cooled to 20-25℃ by the fan and then returns to the circulating water pool to maintain a stable vacuum in the main condenser (actual requirement ≤7kPa); the branch path is diverted from the condenser outlet pipeline to the first heat exchanger, where it exchanges heat with the heat pump working fluid, initially extracting low-grade waste heat at 28-36℃. After heat exchange, the circulating water temperature drops to 25-30℃ and rejoins the main path to return to the cooling tower.

[0021] Under actual unit load variations, this step needs to dynamically adapt to parameter fluctuations: when the unit load drops to 30% (low load condition), the circulating water flow rate decreases by about 10%. At this time, the opening of the inlet valve of the first heat exchanger needs to be reduced to decrease the amount of waste heat extracted and avoid excessive diversion of circulating water leading to insufficient cooling of the condenser; when the unit load rises to 100% (high load condition), the circulating water temperature rises to above 36℃, the opening of the inlet valve of the first heat exchanger needs to be increased to improve the amount of waste heat extracted, and at the same time, all cooling tower fans need to be turned on to ensure the cooling effect of the circulating water and maintain the safe operation of the main unit.

[0022] The heating station is the core link in converting low-grade waste heat into usable thermal energy. The process is designed according to the following steps: initial enhancement of low-temperature waste heat, secondary enhancement of medium-temperature waste heat, and waste heat quality adaptation to the terminal, matching the different temperature requirements of heating network supply and coal-based solid waste fiber production in actual production. In actual operation, the 25-30℃ waste heat working fluid output from the first heat exchanger first enters the low-temperature circulation system, is compressed and heated to 70-80℃ by the low-temperature compressor, and then transfers the heat to the high-temperature circulation working fluid through the intermediate heat exchanger. After absorbing heat, the high-temperature circulation working fluid is further compressed to 120-125℃ by the high-temperature compressor, and then enters the heat pump condenser to exchange heat with the heating network water, heating the heating network water from 40℃ to 75-120℃.

[0023] To meet the specific needs of different end users, this stage requires precise temperature control: when providing hot water at 75-95℃ for the heating network, the high-temperature compressor discharge pressure should be maintained at 1.5-1.7MPa; when providing heat energy at 90-110℃ for the coal-based solid waste fiber forming process in the plant area, the discharge pressure should be increased to 1.7-2.0MPa, and the economizer should be activated to subcool the working fluid, ensuring that the heat pump performance coefficient (COP) remains stable above 1.7 to avoid energy waste. In actual production, if the circulating water temperature drops below 44.5℃ due to unit load fluctuations, the frequency of the low-temperature compressor needs to be increased simultaneously to compensate for insufficient heat source and maintain waste heat to improve efficiency.

[0024] The heating network supply process revolves around adapting to the actual 2.9 million m³ 2 For residential and industrial heating needs, the focus is on ensuring stable water temperature and hydraulic balance. In actual operation, the 75-95℃ primary network hot water output from the heating station is transported to various community heating stations through pipelines. For long-distance transport, a relay pump needs to be started at the midpoint of the pipeline to increase the water pressure to 0.5-1.0MPa to avoid insufficient water supply to distant users. After arriving at the community heating station, the primary network hot water enters a plate heat exchanger to exchange heat with the secondary network return water, heating the secondary network water to 50-60℃ before it is transported to the user end through the secondary network pipeline.

[0025] In actual heating supply at the user end, radiators or underfloor heating in residential buildings receive hot water at 50-60℃, release heat, and then return to the community heating station at 35-40℃. After mixing, the water re-enters the heat exchanger for reheating. To ensure uniform temperature across the entire heating network, electric regulating valves need to be installed in each branch pipe of the primary network. The opening degree is adjusted according to the pressure difference between distant and nearby users. For example, the branch valve opening for distant users is adjusted to 60%-70%, and for nearby users to 40%-50%, so that the temperature fluctuation of the secondary network supply water is controlled within ±1℃, meeting the residents' indoor temperature requirement of 18-22℃.

[0026] As an expandable implementation method, the coal-based solid waste fiber-forming process in the plant can also be deeply coupled with waste heat utilization. In actual production, the 90-110℃ waste heat hot water supplied by the power plant's thermal station first enters the hot air heat exchanger, where it exchanges heat with the ambient temperature air supplied by the blower to generate 50-60℃ hot air, which is used to dry the coal-based solid waste raw materials, reducing the moisture content of the raw materials from more than 5% to less than 1%, thus preventing agglomeration in the subsequent melting process. The dried raw materials enter the melting furnace via a conveyor belt. In the actual melting process, in addition to the main heating device providing a high temperature of 1200-1500℃, waste heat steam from the power plant is also connected for auxiliary heating to maintain furnace temperature fluctuations ≤±20℃, ensuring that the raw materials are fully melted. The molten material is then transported to a centrifugal fiber forming machine, where it is spun into primary fibers of 5-20μm at a speed of 1500-2500r / min, and then enters the shaping oven. The heating coils inside the setting oven are circulated with waste hot water at 90-110℃ to heat the primary fibers to 100-150℃, completing the setting process and increasing the fiber strength to over 200MPa. Finally, the fibers are cooled to 30-40℃ by cooling rollers and then transported to the finished product warehouse for storage. In actual production, if fluctuations in the waste heat temperature cause the setting temperature to drop below 90℃, the auxiliary electric heating inside the oven needs to be briefly activated to ensure the fiber setting effect and avoid substandard quality.

[0027] In actual use at the plant, in an embodiment of a 600MW coal-fired unit, after applying the method of this invention, operational data over a heating season (November 15th to March 15th of the following year, approximately 150 days) showed that: the heat pump COP stability improved from ±15% using the traditional method to ±5%, with the average COP maintained at 1.72; the number of times the condenser vacuum exceeded the standard decreased to 0; and the additional heat provided through waste heat recovery was equivalent to saving approximately 2100 tons of standard coal. Compared to the control period using fixed parameters, the waste heat utilization efficiency improved by approximately 32%.

[0028] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A power plant circulating water waste heat utilization adaptive regulation method based on dynamic heat source characteristics, characterized in that, Comprise the following steps: S1, in the circulating water supply and return water pipeline, the top of the condenser, the secondary side of the heat network is deployed distributed detection unit, real-time acquisition circulating water supply and return water temperature and flow, condenser vacuum, heat supply and return water temperature; the acquisition signal is processed according to the working condition: the circulating water flow ± 10% fluctuation signal is removed by 3σ principle, the heat network temperature is less than or equal to 30s missing signal is filled with interpolation of similar load, after processing, through industrial bus transmission to central controller; S2: based on GRU neural network, a load-circulating water parameter correlation prediction model is constructed, the linkage signal of unit load-circulating water temperature-flow in S1 is input, the fluctuation trend of circulating water temperature and flow in the future 15-30min is predicted; when the predicted temperature is less than 46℃ or the flow reduction is more than 8%, the fluctuation early warning signal is output, and the heat pump pre-adjustment logic is triggered in advance; S3: combined with the main machine safety of vacuum degree less than or equal to 7kPa, heat pump efficiency, terminal demand, through operation to generate adaptive instructions: Heat pump regulation: when the circulating water temperature is 48-55℃, control the heat pump solution pump speed 3000-3500r / min, refrigerant pump flow 360-400m 3 / h; when the temperature is 44.5-48℃, the output speed is increased to 3500-4500r / min, the flow is increased to 400-500m 3 / h, maintain the heat pump efficiency COP≥1.7; Host protection: when the vacuum degree is greater than or equal to 6.5kPa, reduce the heat pump heat extraction amount by 5%-10%, output signal to improve the cooling tower fan speed, and cut off 30% heat extraction amount when the vacuum degree is greater than or equal to 7kPa; Terminal adaptation: when the heat supply water temperature is less than 75℃, output signal to adjust the water mixing valve by 2-5%; S4: convert S3 instructions into signals recognizable by actuators: Host side: the opening change step of condenser regulating valve is less than or equal to 5% / s, and the signal delay is less than or equal to 100ms; Heat pump side: solution pump / refrigerant pump adjustment step ≤ 100 r / min, 20 m 3 / h; Terminal side: the response time of heat network water mixing valve is less than or equal to 2s, and the terminal load fluctuation is adapted; S5: working condition feedback closed loop, feedback unit 1s / time collects executed parameters: main machine vacuum degree confirmation less than or equal to 7kPa, heat pump COP confirmation greater than or equal to 1.7, heat network temperature fluctuation confirmation less than or equal to ± 1℃, generate deviation signal back to central controller.

2. The method of claim 1, wherein, After step S5, S6, working condition iterative optimization, daily supplement variable load regulation and actual field effect correlation data to database, retrain prediction model within 24h, update circulating water temperature prediction threshold under different loads, heat pump and main machine linkage parameters, so that the regulation logic continuously adapts to the variable load working condition of unit.

3. The method of claim 1, wherein, In step S1, the deployment of the detection unit adapts to the working condition demand of circulating water pipeline sensor to avoid turbulent flow area, and the condenser vacuum degree sensor is installed vertically.

4. The method of claim 1, wherein, In step S2, the training logic of the prediction model adapts to the variable load characteristics: according to the low load of 30%-50% of the unit, the medium load of 50%-80%, and the high load of 80%-100%, the working condition interval is divided, and each interval is independently trained sub-model.

5. The method of claim 1, wherein, In step S3, the operation logic of heat pump regulation is to establish the correlation equation among circulating water temperature, pump speed and COP, so that COP is stable at 1.7±0.1 under different temperatures.