A method for automatically controlling the main steam temperature of a power boiler

Abstract

This invention discloses a method for controlling the main steam temperature of a power generation boiler. The method includes collecting step response data of the main steam temperature of the power generation boiler, fitting the data using the least squares method, and establishing a mathematical model of the inertia region and lead-forward region of the main steam control channel of the power generation boiler. Then, a predictive compensation controller is introduced in the parallel feedback loop, and the lag component is separated as a separate term using a constant separation method. Next, the PID parameters are optimized using an artificial fish swarm algorithm, and the optimal solution is substituted into the original parameters. Smith predictive compensation control effectively eliminates the lag effect in the system model, shortens the system settling time, and the artificial fish swarm algorithm significantly improves the controller accuracy, effectively shortens the time for the system to reach the target value, reduces the system overshoot, and minimizes the impact of external disturbances.

Description

Technical Field

[0001] This invention belongs to the field of automatic control technology for power generation boilers, specifically relating to an automatic control method for the main steam temperature of a power generation boiler based on Smith-PID optimization using an artificial fish swarm algorithm. Background Technology

[0002] Boiler main steam temperature is one of the key parameters in boiler control, and its control effectiveness directly affects steam quality and the economical operation of the unit. Boiler main steam temperature is a complex controlled object with significant lag. Changes in combustion, flue gas, manually adjusted desuperheating water volume, external main steam load, and primary / secondary air all influence the main steam temperature. To address these factors, main steam temperature control requires both rapid response to external conditions and a high rate of recovery, as well as strong suppression of changes in desuperheating pressure, ensuring system stability and high-quality steam. However, current conventional DCS software algorithms are modular, with control loops primarily using traditional PID control or cascade PID control, incorporating feedforward signals such as load and fuel quantity. Despite this, for main steam temperature control with significant lag, the actual dynamic performance still exhibits large deviations, long transition times, and dynamic fluctuations, failing to meet high-precision control requirements.

[0003] While intelligent control methods such as fuzzy control, neural network control, sliding mode control, and non-identifying predictive control have achieved good results in main steam temperature control research, their computational complexity and the limitations of the DCS platform prevent their implementation within a DCS system. Implementation using an external APC-controlled computer significantly increases costs. Traditional main steam temperature control for power generation boilers employs PID control (proportional, integral, and derivative control), which linearly combines the proportional (P), integral (I), and derivative (D) values ​​of the deviation to form the control quantity, hence the name PID controller. Traditional PID control achieves good control results for determinate systems, but its performance is poor for systems difficult to describe mathematically precisely. It cannot automatically adjust parameters to provide better control in response to changes in the controlled object. However, with technological advancements, many areas where traditional PID control systems cannot perform well or are even unsuitable for application have emerged. In particular, it cannot effectively regulate the lag caused by the mechanical actions of various stages in the power generation boiler process, and PID control suffers from poor accuracy and long settling times. Summary of the Invention

[0004] To address the numerous uncertainties in the main steam temperature control of industrial power generation boilers, which lead to nonlinearity and significant lag in the control process, this invention provides a method for controlling the main steam temperature of power generation boilers based on an artificial fish swarm algorithm-optimized Smith-PID control. This method improves the control rate by using an artificial fish swarm algorithm to optimize the Smith predictive compensation PID control, thereby solving the technical challenge of precise temperature control.

[0005] The technical solution of the present invention:

[0006] An automatic control method for the main steam temperature of a power generation boiler includes the following steps:

[0007] Step 1: Establish the relevant mathematical model for the inertia zone and guide zone of the main steam control channel of the power generation boiler. The specific steps are as follows:

[0008] By collecting step response data of the main steam of the power generation boiler, and using the least squares method for data fitting, the inertia zone and guide zone models of the main steam control channel of the power generation boiler were obtained.

[0009]

[0010]

[0011] in, This indicates the temperature of the main steam passing through the superheater. This indicates the steam temperature at the desuperheater outlet. Indicates the flow rate of the cooling water. K 1. K 2 represents the gain coefficients of the transfer function in the inert region and the lead region, respectively. T 1. T 2 represents the time constant of the transfer function in the inertial region and the lead-in region, respectively; the initial PID parameters Kp0, Ki0, and Kd0 are calculated using the critical proportional gain method.

[0012] Step 2: Introduce a predictive compensation controller in the parallel feedback loop, and then use the constant separation method to isolate the lag component as a separate term, designing the controller according to conventional methods. .generally This is the transfer function for PID control. Its equivalent transfer function is:

[0013]

[0014] in For the Smith compensation function, from the above equation, we need to...

[0015]

[0016] The compensated system transfer function is

[0017]

[0018] After Smith's forecast compensation, the lag component is separated as a separate item, and its destabilizing effect on the system will be significantly reduced.

[0019] Step 3: Optimize the PID parameters using the artificial fish swarm algorithm, substitute the optimal solution into the original parameters, and run the algorithm. The specific steps are as follows:

[0020] Step 3.1: Parameter Initialization. This involves setting the scale of the artificial fish colony. Fish crowding The field of vision that artificial fish can search Movement step size Number of iterations Configure parameters such as...

[0021] Suppose there are N artificial fish within a certain body of water, and the state of each artificial fish is represented by a vector. It means that among them , The solution vector for X is represented by the food concentration for each artificial fish. It means that, among them, The objective function is represented by the distance between two adjacent artificial fish. This indicates the fish density or crowding level within the water area. This indicates the field of vision that artificial fish can search. Indicates the step size for each movement. The number of trials or iterations during foraging is represented by... express.

[0022] Step 3.2: Calculate the food concentration Y for each artificial fish, i.e., calculate the fitness value based on the objective function. By comparing the fitness values, select the optimal state value and record it on the bulletin board;

[0023] Assume the current position of the artificial fish is It randomly selects a swimming location within the water area it perceives. Therefore, the mathematical model for the swimming of artificial fish can be expressed as:

[0024]

[0025] In the formula: —Random function, and .

[0026] Simultaneously assuming position Food concentration at , Food concentration at If the food concentration Greater than Then the artificial fish will be from the location To position Move one step, step length Meanwhile, the current value of the artificial fish's next search can be represented as:

[0027]

[0028] Conversely, if the food concentration Less than Then the artificial fish will find a new location within its field of vision. Then compare the food concentration at the two locations. If foraging is ongoing, the number of iterations... If the artificial fish reaches its maximum food concentration but still cannot find a location with a higher concentration, it will randomly move one step within its field of vision. Its mathematical expression is:

[0029]

[0030] Step 3.3: Execute aggregation and tail-chasing behaviors, with foraging behavior as the default behavior. Select the optimal behavior to find food based on the fitness value calculated in step (2);

[0031] Assume the artificial fish is currently located at... The number of fish within its visible range is used Indicates location The food concentration at that location is So, the center of the fish school The mathematical expression at the location is:

[0032]

[0033] Assuming the center of the fish school is located The food concentration at that location is Whether artificial fish move toward the center of the fish school can be determined by the following formula.

[0034]

[0035] If the above formula holds true, it means that there is sufficient food at the center of the fish school and the school is not overcrowded. In this case, the artificially created fish will move towards the center of the school, forming a swarming behavior, the mathematical expression of which is:

[0036]

[0037] Conversely, the artificial fish will then perform step (2).

[0038] Assume the current position of the artificial fish is Within its field of vision, the highest concentration of food that can be detected ( The position corresponding to ) is So, will artificial fish move to a certain location? The movement can be determined by the following formula.

[0039]

[0040] If the above formula is true, it indicates the location. If there is ample food and the fish population is not overcrowded, then the artificial fish will move to the location where... The mathematical expression for movement is:

[0041]

[0042] Conversely, the artificial fish will perform step (2).

[0043] The beneficial effects of this invention are as follows: This invention is based on the artificial fish swarm algorithm to optimize the main steam temperature control method of the power generation boiler using Smith-PID. By using Smith predictive compensation control, the lag effect in the system model can be effectively eliminated, the system adjustment time can be shortened, and the artificial fish swarm algorithm can be used to optimize the controller parameters, which can significantly improve the accuracy of the controller, effectively shorten the time for the system to reach the target value, reduce the overshoot of the system, and reduce the influence of external disturbances. Attached Figure Description

[0044] Figure 1 This is a process control diagram for the main steam temperature of a power generation boiler.

[0045] Figure 2 This is a process control block diagram for the main steam temperature of a power generation boiler.

[0046] Figure 3 Optimize the flowchart for artificial fish swarms. Detailed Implementation

[0047] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. Example 1

[0048] A method for controlling the main steam temperature of a power generation boiler based on a Smith-PID algorithm optimized by an artificial fish swarm algorithm is described below. Figure 1 As shown, the method is characterized by the following steps:

[0049] Step 1: Establish the relevant mathematical model for the inertia zone and guide zone of the main steam control channel of the power generation boiler. The specific steps are as follows:

[0050] like Figure 2 The model shown is obtained by collecting step response data of the main steam of the power generation boiler and fitting the data using the least squares method to obtain the inert zone and guide zone model of the main steam control channel of the power generation boiler.

[0051]

[0052]

[0053] in, This indicates the temperature of the main steam passing through the superheater. This indicates the steam temperature at the desuperheater outlet. This represents the flow rate of the desuperheating water. The initial PID parameters Kp0, Ki0, and Kd0 are calculated using the critical proportional gain method.

[0054] Step 2: Introduce a predictive compensation controller in the parallel feedback loop, and then use the constant separation method to isolate the lag component as a separate term, designing the controller according to conventional methods. .generally This is the transfer function for PID control. Its equivalent transfer function is:

[0055]

[0056] in For the Smith compensation function, from the above equation, we need to...

[0057]

[0058] The compensated system transfer function is

[0059]

[0060] After Smith's forecast compensation, the lag component is separated as a separate item, and its destabilizing effect on the system will be significantly reduced.

[0061] Step 3: Optimize the PID parameters using the artificial fish swarm algorithm, substitute the optimal solution into the original parameters, and run the algorithm. The specific steps are as follows:

[0062] Step 3.1: Parameter Initialization. This involves setting the scale of the artificial fish colony. Fish crowding The field of vision that artificial fish can search Movement step size Number of iterations Configure parameters such as...

[0063] Suppose there are N artificial fish within a certain body of water, and the state of each artificial fish is represented by a vector. It means that among them , The solution vector for X is represented by the food concentration for each artificial fish. It means that, among them, The objective function is represented by the distance between two adjacent artificial fish. This indicates the fish density or crowding level within the water area. This indicates the field of vision that artificial fish can search. Indicates the step size for each movement. The number of trials or iterations during foraging is represented by... express.

[0064] Step 3.2: Calculate the food concentration Y for each artificial fish, i.e., calculate the fitness value based on the objective function. By comparing the fitness values, select the optimal state value and record it on the bulletin board;

[0065] Assume the current position of the artificial fish is It randomly selects a swimming location within the water area it perceives. Therefore, the mathematical model for the swimming of artificial fish can be expressed as:

[0066]

[0067] In the formula: —Random function, and .

[0068] Simultaneously assuming position Food concentration at , Food concentration at If the food concentration Greater than Then the artificial fish will be from the location To position Move one step, step length Meanwhile, the current value of the artificial fish's next search can be represented as:

[0069]

[0070] Conversely, if the food concentration Less than Then the artificial fish will find a new location within its field of vision. Then compare the food concentration at the two locations. If foraging is ongoing, the number of iterations... If the artificial fish reaches its maximum food concentration but still cannot find a location with a higher concentration, it will randomly move one step within its field of vision. Its mathematical expression is:

[0071]

[0072] Step 3.3: Execute aggregation and tail-chasing behaviors, with foraging behavior as the default behavior. Select the optimal behavior to find food based on the fitness value calculated in step (2);

[0073] Assume the artificial fish is currently located at... The number of fish within its visible range is used Indicates location The food concentration at that location is So, the center of the fish school The mathematical expression at the location is:

[0074]

[0075] Assuming the center of the fish school is located The food concentration at that location is Whether artificial fish move toward the center of the fish school can be determined by the following formula.

[0076]

[0077] If the above formula holds true, it means that there is sufficient food at the center of the fish school and the school is not overcrowded. In this case, the artificially created fish will move towards the center of the school, forming a swarming behavior, the mathematical expression of which is:

[0078]

[0079] Conversely, the artificial fish will then perform step (2).

[0080] Assume the current position of the artificial fish is Within its field of vision, the highest concentration of food that can be detected ( The position corresponding to ) is So, will artificial fish move to a certain location? The movement can be determined by the following formula.

[0081]

[0082] If the above formula is true, it indicates the location. If there is ample food and the fish population is not overcrowded, then the artificial fish will move to the location where... The mathematical expression for movement is:

[0083]

[0084] Conversely, the artificial fish will perform step (2).

[0085] Artificial fish swarm optimization process as follows Figure 3As shown.

[0086] A method for controlling the main steam temperature of a power generation boiler based on a Smith-PID algorithm optimized by an artificial fish swarm algorithm is disclosed. The system includes a Smith control module, a gain module, a PID control module, a control valve, an optimization module, and a temperature transmitter module. The outputs of the Smith control module and the optimization module are connected to the input of the gain module. The output of the gain module is connected to the input of the PID control module. The output of the PID control module is connected to the control valve. The control valve outputs a temperature to the controlled object, adjusting the main steam outlet temperature of the boiler. The main steam outlet temperature is fed back to the input temperature setpoint via the temperature transmitter module.

[0087] The overall process has been optimized as follows:

[0088] (1) Establish a first-order inertial plus pure time delay model for the main steam temperature of the boiler;

[0089] (2) Eliminate the lag effect by introducing a predictive compensation controller in the parallel feedback loop;

[0090] (3) The artificial fish swarm algorithm is used to optimize the parameters and then substituted into the controller parameters, as follows:

[0091] 3.1 Parameter Initialization. This section discusses the initialization of parameters for the artificial fish swarm size. Fish crowding The field of vision that artificial fish can search Movement step size Number of iterations Configure parameters such as...

[0092] 3.2 Calculate the food concentration value for each artificial fish, and then calculate the fitness value based on the objective function. By comparing the fitness values, select the optimal state value and record it on the bulletin board;

[0093] 3.3 Perform aggregation and tail-chasing behaviors, with foraging behavior as the default behavior. Select the optimal behavior to find food based on the fitness value calculated in step (2);

[0094] 3.4 Perform the selected action in step (3) and update the food concentration value;

[0095] 3.5 Compare the updated food concentration values ​​with the values ​​on the bulletin board, and record the optimal value on the bulletin board.

[0096] 3.6 Determine if the termination condition is met. The termination condition is generally that the fitness value reaches the minimum error or the maximum number of iterations. If the termination condition is met, output the optimal value and complete the optimization. Otherwise, repeat step (3) until the termination condition is met.

Claims

1. An automatic control method for the main steam temperature of a power generation boiler, characterized in that, Includes the following steps: Step 1: Establish relevant mathematical models for the inertia zone and guide zone of the main steam control channel of the power generation boiler; by collecting step response data of the main steam of the power generation boiler, the data is fitted using the least squares method to obtain the models of the inertia zone and guide zone of the main steam control channel of the power generation boiler: ； ； in, This indicates the temperature of the main steam passing through the superheater. This indicates the steam temperature at the desuperheater outlet. Indicates the flow rate of the cooling water. K 1. K 2 represents the gain coefficients of the transfer function in the inert region and the lead region, respectively. T 1. T 2 represents the time constant of the transfer function in the inertial region and the lead-out region, respectively; the initial PID parameters Kp0, Ki0, and Kd0 are calculated using the critical proportional gain method; Step 2: By introducing a predictive compensation controller in the parallel feedback loop, the lag component is separated as a separate term using the constant separation method; Step 3: Optimize the PID parameters using the artificial fish swarm algorithm, and substitute the optimal solution into the original parameters; the specific steps are as follows: Step 3.1: Parameter initialization, setting the artificial fish swarm size. Fish crowding The field of vision that artificial fish can search Movement step size Number of iterations Configure the parameters; Suppose there are N artificial fish within a certain body of water, and the state of each artificial fish is represented by a vector. It means that among them , The solution vector for X is represented by the food concentration for each artificial fish. It means that, among them, The objective function is represented by the distance between two adjacent artificial fish. This indicates the fish density or crowding level within the water area. This indicates the field of vision that artificial fish can search. Indicates the step size for each movement. The number of trials or iterations during foraging is represented by... express; Step 3.2: Calculate the food concentration Y for each artificial fish, that is, calculate the fitness value according to the objective function, select the optimal state value by comparing the fitness values, and record it on the bulletin board; Assume the current position of the artificial fish is It randomly selects a swimming location within the water area it perceives. Therefore, the mathematical model for the swimming of artificial fish can be expressed as: ； In the formula: —Random function, and Simultaneously, assume the location Food concentration at , Food concentration at If the food concentration Greater than Then the artificial fish will be from the location To position Move one step, step length Meanwhile, the current value for the next search by the artificial fish is represented as: ； Conversely, if the food concentration Less than Then the artificial fish will find a new location within its field of vision. Then compare the food concentration at the two locations; if foraging is in progress, the number of iterations... If the artificial fish reaches its maximum food concentration but still cannot find a location with a higher concentration, it will randomly move one step within its field of vision. Its mathematical expression is: ； Step 3.3: Execute the gathering and tailing behaviors, where the foraging behavior is the default behavior; select the optimal behavior to find food based on the fitness value calculated in step (2); Assume the artificial fish is currently located at... The number of fish within its visible range is used Indicates location The food concentration at that location is So, the center of the fish school The mathematical expression at the location is: ； Assuming the center of the fish school is located The food concentration at that location is Whether the artificial fish moves toward the center of the fish school can be determined by the following formula; ； If the above formula holds true, it means that there is sufficient food at the center of the fish school and the school is not overcrowded. In this case, the artificially created fish will move towards the center of the school, forming a swarming behavior, the mathematical expression of which is: ； Conversely, the artificial fish will then perform step (2); assuming the artificial fish's current position is... Within its field of vision, the highest concentration of food that can be detected ( The position corresponding to ) is So, will artificial fish move to a certain location? The movement is determined by the following formula: ； If the above formula is true, it indicates the location. If there is ample food and the fish population is not overcrowded, then the artificial fish will move to the location where... The mathematical expression for movement is: ； Conversely, the artificial fish will perform step (2).

2. The automatic control method for main steam temperature of a power generation boiler according to claim 1, characterized in that: In step 2, the controller is designed. , The transfer function for PID control is given by: ； in For the Smith compensation function, from the above equation, we need to... ； The compensated system transfer function is ； After Smith's forecast compensation, the lag component is separated as a separate item, and its destabilizing effect on the system will be significantly reduced.

Images