A deep start-stop peak shaving whole-process intelligent optimization control method for a thermal power unit

Abstract

This invention relates to the field of thermal power generation technology and discloses a method for intelligent optimization control of the entire process of deep start-up and peak shaving of thermal power units. This invention collects key state parameters of the unit in real time and identifies the stage the unit is in based on these parameters. Before start-up and shutdown, it matches the optimal start-up and shutdown target path from a historical optimal control database based on the unit's initial state vector. During start-up and shutdown, it manages and controls key indicators of the unit, sets safety boundary conditions for each stage, and automatically pauses and optimizes the start-up and shutdown target path when constraints are exceeded. Simultaneously, based on the key state data of the unit during start-up and shutdown, it automatically adjusts the target rates of key components, thereby improving the intelligence of unit start-up and shutdown control.

Description

Technical Field

[0001] This invention relates to the field of thermal power generation technology, specifically to an intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units. Background Technology

[0002] With the rapid growth of installed capacity of new energy power generation such as wind and solar power, the demand of the power grid for peak shaving of thermal power units is increasing. To absorb new energy sources, thermal power units need to frequently perform deep peak shaving, and even daily start-up and shutdown. Traditional start-up and shutdown processes for thermal power units typically employ a combination of manual operation by operators and sequential control by a conventional distributed control system (DCS), which presents the following technical problems:

[0003] Reliance on human experience: The control of key parameters such as the heating rate, pressure rate, and turbine acceleration rate during start-up and shutdown relies heavily on the experience of operators. Significant differences in operation among different personnel make it difficult to guarantee the safety and economy of each start-up and shutdown. Long start-up and shutdown times and high energy consumption: For safety reasons, operators often adopt conservative operating strategies, resulting in long start-up and shutdown times and high fuel and plant power consumption. Difficulty in real-time avoidance of safety risks: Traditional control methods lack dynamic closed-loop monitoring of safety boundaries such as temperature change rate, pressure change rate, and bearing vibration. Once the rate exceeds the limit or the stress exceeds the limit, it is difficult to automatically correct the deviation in time, easily causing equipment fatigue damage or even accidents. Lack of self-learning and path optimization capabilities: Existing sequential control systems can only execute according to preset logical steps and cannot intelligently match the optimal start-up and shutdown path based on the current initial state of the unit (such as turbine metal temperature, steam drum pressure, ambient temperature, etc.), nor can they learn from the actual process of each start-up and shutdown and improve subsequent target paths. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an intelligent optimization control method for the entire process of deep start-up and peak shaving in thermal power units. This method has the advantages of real-time acquisition of key status parameters of the unit, matching the optimal start-up and shutdown target path before start-up and shutdown, controlling key indicators during start-up and shutdown, judging whether limits are exceeded based on constraints, and automatically adjusting the target speed of key components during start-up and shutdown, thus solving the aforementioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for intelligent optimization control of the entire process of deep start-up and peak shaving in thermal power units, comprising the following steps:

[0006] S1: Real-time acquisition of key status parameters of the unit, and identification of the unit's current status based on these parameters.

[0007] stage;

[0008] S2: Before start-up and shutdown, the optimal start-up and shutdown target path is matched from the historical best control database based on the unit's initial state vector;

[0009] S3: Control key indicators during start-up and shutdown;

[0010] S4: Set safety boundary conditions for each stage. When the constraint conditions are exceeded, automatically pause and optimize the start and stop target path.

[0011] S5: Based on the unit's critical status data, automatically adjust the target rate of key components.

[0012] As a preferred embodiment of the present invention, the S2 step of matching the optimal start / stop target path includes the following steps:

[0013] S2.1: Calculate the Euclidean distance similarity index for each start / stop path in the historical best control database; S2.2: Select the Euclidean distance similarity index. The historical start-stop path will be used as the target path for this start-stop operation.

[0015] As a preferred technical solution of the present invention, the Euclidean distance similarity index The expression is as follows:

[0016] ;

[0017] in, This represents the weighted Euclidean distance between the current state of the unit and the initial state of its historical start-stop paths; The first indicates the current status of the unit. Each state component For the steam turbine metal temperature, For the steam drum pressure, The ambient temperature; Indicates the first in the historical start-stop path One state component; Indicates the first The normal range of variation for each state component; This represents the summation operation; Indicates the first The weighting coefficients of each state component.

[0018] As a preferred technical solution of the present invention, step S3 controls key indicators, including the following steps:

[0019] S3.1: Calculate the deviation between the actual value and the target value;

[0020] S3.2: The proportional-integral method is used to correct the control quantity, and its expression is as follows:

[0021] ;

[0022] in, Indicates time The output value of the control quantity; Indicates time The deviation between the actual values ​​of the unit's key indicators and the target values ​​in the target start-stop path; Indicates the proportional gain coefficient; Indicates the integral gain coefficient; Represents a time variable.

[0023] As a preferred embodiment of the present invention, the relevant expression for the deviation between the actual value and the target value in S3.1 is as follows:

[0024] ;

[0025] in, Indicates time The deviation between the actual values ​​of the unit's key indicators and the target values ​​in the target start-stop path; Indicates time Key performance indicators (KPIs) are the target values ​​in the start-stop path. Indicates time The actual value of the key indicator.

[0026] As a preferred embodiment of the present invention, the S4 constraint condition is as follows:

[0027] ;

[0028] ;

[0029] ;

[0030] in, Indicates time The rate of heating; Represents absolute value; Indicates the maximum permissible rate of temperature increase; Indicates time The rate of change of pressure; Indicates the maximum permissible rate of pressure change; Indicates time The bearing vibration amplitude; This indicates the bearing vibration limit.

[0031] As a preferred embodiment of the present invention, the relevant expression for optimizing the start / stop target path is as follows:

[0032] ;

[0033] in, Indicates time Optimized values ​​of key indicators in the current start-stop target path; Indicates time The original values ​​of key indicators in the current start / stop target path; Indicates time The actual values ​​of the key performance indicators currently being executed by the unit; It is a relaxation factor.

[0034] As a preferred embodiment of the present invention, S5 automatically adjusts the target rate of key components based on the key status data of the unit, including the following steps:

[0035] S5.1: Real-time wear rate of key components of computer group equipment, expressed as follows:

[0036] ;

[0037] in, Indicates the critical component at a given time. Loss rate; Indicates the critical component at a given time. Thermal stress; This indicates the maximum allowable stress for this component over a long period of time; Indicates the instantaneous ultimate stress;

[0038] S5.2: Real-time loss rate of critical components When the set danger threshold is reached, the target rate of the component is updated.

[0039] As a preferred embodiment of the present invention, the relevant expression for updating the target rate of the component in step S5.2 is as follows:

[0040] ;

[0041] in, This represents the updated value of the component's target rate. Indicates the attenuation factor; This represents the initial value of the target speed of the component.

[0042] As a preferred technical solution of the present invention, the attenuation factor The expression is as follows:

[0043] ;

[0044] in, Indicates the critical component at a given time. Loss rate; Indicates the maximum allowable loss rate; This indicates taking the maximum value.

[0045] Compared with existing technologies, this invention provides an intelligent optimization control method for the entire process of deep start-up and peak shaving in thermal power units, which has the following beneficial effects:

[0046] This invention identifies the unit's key status parameters in real time and based on these parameters...

[0047] At each stage of the unit's operation, before the unit starts or stops, the optimal start-up and shutdown target path is matched from the historical best control database based on the unit's initial state vector. During the start-up and shutdown process, key indicators of the unit are controlled and safety boundary conditions are set for each stage. When the constraint conditions are detected to exceed the limit, the start-up and shutdown target path is automatically suspended and optimized. At the same time, based on the key state data of the unit during the start-up and shutdown process, the target rate of key components is automatically adjusted, which improves the intelligence of the unit's start-up and shutdown control. Attached Figure Description

[0048] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] Please see Figure 1 A method for intelligent optimization control of the entire process of deep start-up and peak shaving of thermal power units includes the following steps:

[0051] S1: Real-time acquisition of key status parameters of the unit, and identification of the stage of the unit based on key status parameters. The unit is divided into several key stages from shutdown to grid connection and power generation (start-up process) and from full load to safe shutdown (stop process), including boiler water filling stage, ignition and heating stage, turbine start-up stage, grid connection and load-bearing stage, load reduction and disconnection stage, shutdown and cooling stage, etc.

[0052] S2: Before start-up and shutdown, the optimal start-up and shutdown target path is matched from the historical best control database based on the unit's initial state vector, including the following steps:

[0053] S2.1: Calculate the Euclidean distance similarity index for each start / stop path in the historical best control database.

[0054] Its expression is as follows:

[0055] ;

[0056] in, This represents the weighted Euclidean distance between the current state of the unit and the initial state of its historical start-stop paths; The first indicates the current status of the unit. Each state component For the steam turbine metal temperature, For the steam drum pressure, The ambient temperature; Indicates the first in the historical start-stop path One state component; Indicates the first The normal range of variation for each state component; This represents the summation operation; Indicates the first The weighting coefficients of each state component;

[0057] During the deep start-stop peak shaving process of thermal power units, the initial states (such as turbine metal temperature, steam drum pressure, and ambient temperature) are different for each start-stop. If a fixed start-stop path is used, it cannot adapt to the state differences, which may lead to safety risks or increased time and energy consumption. The design principle of this step is: based on the similarity matching principle, the current unit's state vector is compared with the initial state vectors of each successful start-stop path in the historical database using a weighted Euclidean distance calculation. By assigning weights to different state components (for example, turbine metal temperature has the highest weight because it directly affects thermal stress), the similarity between the current state and the historical path can be quantified. Finally, the historical path with the smallest distance is selected as the target path for this start-stop, thereby achieving intelligent matching for path reuse when the states are similar, avoiding the need to replan the path for each start-stop, and improving the adaptability and reliability of the control.

[0058] In this step, the weighted Euclidean distance The magnitude of directly reflects the comprehensive difference between the current state of the thermal power unit and the initial state of a certain historical start-up and shutdown path; where This represents the deviation between the current value and historical values, divided by... Deviations of different dimensions (°C, MPa, °C) are made dimensionless so that they can be linearly added; weighting This reflects the different degrees of influence of various state variables on the start-up and shutdown process: turbine metal temperature ( The weight of steam drum pressure is relatively large because it determines thermal stress and lifespan loss; Affects boiler energy storage and pressure rise rate; ambient temperature ( This affects heat dissipation and combustion regulation; the final calculated minimum... The historical path corresponding to the value physically means that when a certain start-up or shutdown began in the past, the unit's thermal state, pressure state, and environmental conditions were most similar to the current state. Therefore, the rate, time, and operating steps adopted by this path are most suitable for the current operating conditions.

[0059] S2.2: Select Euclidean distance similarity index The historical start-stop path is used as the target path for this start-stop;

[0061] S3: During the start-up and shutdown process, key indicators are controlled, including the following steps:

[0062] S3.1: The deviation between the actual value and the target value is calculated, and its expression is as follows:

[0063] ;

[0064] in, Indicates time The deviation between the actual values ​​of the unit's key indicators and the target values ​​in the target start-stop path; Indicates time Key performance indicators (KPIs) are the target values ​​in the start-stop path. Indicates time The actual values ​​of key indicators;

[0065] During the deep start-up and peak-shaving process of thermal power units, key indicators such as heating rate, pressure rise rate, turbine rise rate, and load change rate need to follow the predetermined target start-up and shutdown path in real time. To achieve this following target, it is necessary to first quantify the degree of difference between the actual value and the expected value. The design principle of this step is: defining the deviation amount. For target value Subtract the actual value This deviation serves as the input signal for the subsequent closed-loop controller; when the actual value is lower than the target value... If the value is greater than 0, the controller will increase the control input (such as increasing the fuel command or opening the valve wider) to improve the actual value; conversely, if the actual value is greater than the target value, the controller will increase the control input (such as increasing the fuel command or opening the valve wider) to improve the actual value; If the value is less than 0, the controller will reduce the control input to lower the actual value. By calculating the deviation in real time and adjusting the control action according to the magnitude of the deviation, it can ensure that the actual start-stop process continuously approaches the target path, thereby achieving precise closed-loop control of the entire process.

[0066] S3.2: The proportional-integral method is used to correct the control quantity, and its expression is as follows:

[0067] ;

[0068] in, Indicates time The output value of the control quantity; Indicates time The deviation between the actual values ​​of the unit's key indicators and the target values ​​in the target start-stop path; Indicates the proportional gain coefficient; Indicates the integral gain coefficient; Represents a time variable;

[0069] During the start-up and shutdown of thermal power units, only the instantaneous deviation at the current moment is considered. While proportional control (considering only the current error) offers a fast response, it cannot eliminate accumulated static errors, potentially causing the actual value to deviate from the target value over a long period. Therefore, this step introduces an integral term into the proportional control, forming a proportional-integral controller. Its design principle is: control output... It consists of two superimposed parts: Part 1 Provide an immediate response to the current deviation to quickly correct transient deviations; Part Two The integral term continuously adjusts the control quantity as long as the historical deviation is not zero, based on the cumulative integral of all past deviations, until the deviation is zero in steady state. This combination method takes into account both dynamic response speed and steady-state error-free characteristics, and is suitable for the control of thermodynamic processes with large inertia, such as heating rate and pressure rate.

[0070] S4: Set safety boundary conditions for each stage. When the constraint conditions are exceeded, automatically pause and optimize the start and stop target path.

[0071] The constraints are as follows:

[0072] ;

[0073] ;

[0074] ;

[0075] in, Indicates time The rate of heating; Represents absolute value; Indicates the maximum permissible rate of temperature increase; Indicates time The rate of change of pressure; Indicates the maximum permissible rate of pressure change; Indicates time The bearing vibration amplitude; Indicates the bearing vibration limit;

[0076] The relevant expressions for optimizing the start / stop target path are as follows:

[0077] ;

[0078] in, Indicates time Optimized values ​​of key indicators in the current start-stop target path; Indicates time The original values ​​of key indicators in the current start / stop target path; Indicates time The actual values ​​of the key performance indicators currently being executed by the unit; It is a relaxation factor;

[0079] In this formula, the difference Reflecting at any moment Deviation between actual unit performance and original settings: A positive value indicates that the actual value is higher than the original target (e.g., the actual temperature is higher than the target temperature); a negative value indicates that the actual value is lower than the original target; relaxation factor. This determines the proportion of deviations that are incorporated into the new path; physically, Equivalent to a learning rate, relatively small (e.g., 0.2) means that the adjustment to the original path is relatively conservative, prioritizing historical experience; a larger value... (e.g., 0.8) means greater trust in the actual operational data, enabling faster correction of unreasonable parts of the original path; by selecting appropriate... New Path Physically, this can be interpreted as: an optimized trajectory that compromises between theoretical expectations and practical feasibility, which retains the safety framework of the original path while incorporating better operating points discovered in actual execution, thereby making the subsequent start-up and shutdown process more in line with the actual dynamic characteristics of the unit.

[0080] S5: Based on the unit's critical status data, automatically adjust the target rate of critical components, including the following steps:

[0081] S5.1: Real-time wear rate of key components of computer group equipment, expressed as follows:

[0082] ;

[0083] in, Indicates the critical component at a given time. Loss rate; Indicates the critical component at a given time. Thermal stress; This indicates the maximum allowable stress for this component over a long period of time; Indicates the instantaneous ultimate stress;

[0084] Loss rate The magnitude of this value directly represents the degree to which thermal stress consumes the equipment's lifespan at the current moment. =0 indicates that the stress is exactly equal to the upper limit of the allowable long-term operation, and the lifespan wear rate is at a normal level; if =0.5 indicates that the stress has reached the midpoint between the safe zone and the limit zone. If the same heating rate is maintained at this point, lifespan loss will accelerate. A value close to 1 indicates that the unit is in an extremely high-risk state and the rate must be reduced immediately or the start-up and shutdown process must be suspended.

[0085] S5.2: Real-time loss rate of critical components When the set danger threshold is reached, the target rate of the component is updated, and the relevant expression is as follows:

[0086] ;

[0087] in, This represents the updated value of the component's target rate. Indicates the attenuation factor; The initial value representing the target speed of the component;

[0088] During the start-up and shutdown of thermal power units, when the real-time wear rate of key components... When the preset danger threshold is exceeded, it means that the current thermal stress has significantly deviated from the safe zone. If the original target rate continues, further deterioration will occur. Execution may accelerate low-cycle fatigue damage and even cause equipment failure; the design principle of this step is: to introduce a factor based on attenuation. The rate correction mechanism directly multiplies the original target rate by... To obtain a new, lower target rate ;

[0089] Attenuation factor The expression is as follows:

[0090] ;

[0091] in, Indicates the critical component at a given time. Loss rate; Indicates the maximum allowable loss rate;

[0092] During the start-up and shutdown of thermal power units, when the real-time wear rate of key components... When increasing the peak load, the target rate needs to be appropriately reduced to protect the equipment's lifespan. However, the rate reduction should not be too large to avoid excessively long start-up and shutdown times that could affect peak load response capabilities. The design principle of this step is to construct an attenuation factor. This ensures that the loss rate approaches 1 when it is low (without deceleration), and approaches the maximum allowable value when the loss rate is close to the maximum allowable value. The speed tends towards 0.5 (at most half the speed); specifically, a linear decreasing relationship is adopted. and use The function's lower bound is set to 0.5 to prevent excessive deceleration from causing process stalling; this design allows the target rate to be adjusted to... This allows for a more cautious adjustment rate as the loss rate increases, while still maintaining a minimum of 50% of the baseline rate, thus achieving a balance between lifetime protection and peak-shaving economics.

[0093] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent optimization control of the entire process of deep start-up and peak shaving in thermal power units, characterized in that: S1: Real-time acquisition of key status parameters of the unit, and identification of the unit's current status based on these parameters. stage; S2: Before start-up and shutdown, the optimal start-up and shutdown target path is matched from the historical best control database based on the unit's initial state vector; S3: Control key indicators during start-up and shutdown; S4: Set safety boundary conditions for each stage. When the constraint conditions are exceeded, automatically pause and optimize the start and stop target path. S5: Based on the unit's critical status data, automatically adjust the target rate of key components.

2. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 1, characterized in that: The S2 matching of the optimal start / stop target path includes the following steps: S2.1: Calculate the Euclidean distance similarity index for each start / stop path in the historical best control database; S2.2: Select the Euclidean distance similarity index. The historical start-stop path will be used as the target path for this start-stop operation. path.

3. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 2, characterized in that: The Euclidean distance similarity index The expression is as follows: ； in, This represents the weighted Euclidean distance between the current state of the unit and the initial state of its historical start-stop paths; The first indicates the current status of the unit. Each state component For the steam turbine metal temperature, For the steam drum pressure, The ambient temperature; Indicates the first in the historical start-stop path One state component; Indicates the first The normal range of variation for each state component; This represents the summation operation; Indicates the first The weighting coefficients of each state component.

4. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 2, characterized in that: S3 controls key indicators. Includes the following steps: S3.1: Calculate the deviation between the actual value and the target value; S3.2: The proportional-integral method is used to correct the control quantity, and its expression is as follows: ； in, Indicates time The output value of the control quantity; Indicates time The deviation between the actual values ​​of the unit's key indicators and the target values ​​in the target start-stop path; Indicates the proportional gain coefficient; Indicates the integral gain coefficient; Represents a time variable.

5. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 4, characterized in that: The relevant expression for the deviation between the actual value and the target value in S3.1 is as follows: ； in, Indicates time The deviation between the actual values ​​of the unit's key indicators and the target values ​​in the target start-stop path; Indicates time Key performance indicators (KPIs) are the target values ​​in the start-stop path. Indicates time The actual value of the key indicator.

6. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 4, characterized in that: The S4 constraint conditions are as follows: ； ； ； in, Indicates time The rate of heating; Represents absolute value; Indicates the maximum permissible rate of temperature increase; Indicates time The rate of change of pressure; Indicates the maximum permissible rate of pressure change; Indicates time The bearing vibration amplitude; This indicates the bearing vibration limit.

7. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 6, characterized in that: The relevant expressions for optimizing the start / stop target path are as follows: ； in, Indicates time Optimized values ​​of key indicators in the current start-stop target path; Indicates time The original values ​​of key indicators in the current start / stop target path; Indicates time The actual values ​​of the key performance indicators currently being executed by the unit; It is a relaxation factor.

8. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 7, characterized in that: Based on the unit's critical status data, S5 automatically adjusts the target rates of key components. Includes the following steps: S5.1: Real-time wear rate of key components of computer group equipment, expressed as follows: ； in, Indicates the critical component at a given time. Loss rate; Indicates the critical component at a given time. Thermal stress; This indicates the maximum allowable stress for this component over a long period of time; Indicates the instantaneous ultimate stress; S5.2: Real-time loss rate of critical components When the set danger threshold is reached, the target rate of the component is updated.

9. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 8, characterized in that: The relevant expression for updating the target rate of the component in S5.2 is as follows: ； in, This represents the updated value of the component's target rate. Indicates the attenuation factor; This represents the initial value of the target speed of the component.

10. The intelligent optimization control method for the entire process of deep start-up and peak shaving of thermal power units according to claim 9, characterized in that: The attenuation factor The expression is as follows: ； in, Indicates the critical component at a given time. Loss rate; Indicates the maximum allowable loss rate; This indicates taking the maximum value.

Images