Chemical ammonification cascade control method based on improved particle swarm optimization feedforward model

Abstract

The invention discloses a chemical ammonification cascade control method based on an improved particle swarm optimization feedforward model. The method comprises the following steps: establishing the improved particle swarm optimization feedforward model; the feed water flow, the ammonia solution conductivity and the economizer inlet pH set value are input into an improved particle swarm optimization feedforward model, and a first prediction frequency of an ammonia adding pump is generated; the difference value between the actual pH value of the economizer inlet and the pH set value of the economizer inlet is input into an outer ring PI controller, and a pH set value of the deaerator inlet is generated; the difference value of the actual pH value of the inlet of the deaerator and the pH set value of the inlet of the deaerator is input into an inner ring PI controller, and a second predicted frequency of the ammonia adding pump is generated; adding the first predicted frequency of the ammonia adding pump and the second predicted frequency of the ammonia adding pump to obtain a total ammonia adding pump frequency, so as to realize automatic control of chemical ammonia adding; according to the method, the feedforward model structure is designed, the feedforward model parameters are identified by adopting the improved particle swarm optimization, and accurate prediction of the ammonia adding demand quantity can be realized.

Description

Technical Field

[0001] The invention relates to the technical field of chemical automatic control, and in particular to a chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model. Background Art

[0002] The coal-fired unit feedwater ammonia system adds ammonia water to the boiler feedwater pipe to keep the boiler feedwater weakly alkaline and prevent the boiler feedwater from corroding metals. At present, most units manually adjust the frequency of the ammonia pump to change the amount of ammonia added and control the pH value of the feedwater. However, due to the large time lag characteristics of the chemical online instrument measurement sampling link, the change in the pH value of the feedwater lags behind the flow change under variable load conditions, resulting in the operator being unable to add ammonia in a timely and accurate manner, causing the pH value of the feedwater to fluctuate over a wide range, posing a risk of metal corrosion and affecting the safe operation of the unit. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention provides a chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model, which aims to suppress load interference, maintain the stability of the pH value of the feed water, and ensure the safe operation of the unit.

[0004] To achieve the above object, the present invention provides the following technical solution: a chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model, comprising the following steps: Step S1: Establishing an improved particle swarm optimization feedforward model; By screening the variables related to the frequency of the ammonia pump during the operation of the thermal power unit, and considering the time lag relationship between the mechanism of the thermal power unit and the relevant variables, a feedforward model is constructed, and the improved particle swarm algorithm is used to identify the parameters of the feedforward model, and the improved particle swarm optimization feedforward model is obtained; The specific process of using the improved particle swarm algorithm to identify the parameters of the feedforward model is as follows: Step S1.21: Based on the feedforward model parameters , , , , , set the improved particle swarm algorithm The position of each particle is The fitness function of the improved particle swarm algorithm is the prediction error of the frequency of the ammonia pump The variance of is the water flow lag time, is the conductivity lag time of ammonia solution, is the pH lag time at the economizer inlet, and are the weight and bias of the economizer inlet pH value, Indicates time; Step S1.22: Initialize the improved particle swarm algorithm; determine the number of individual particles, the initial position and speed of each individual particle; Step S1.23: Calculate the fitness of each individual particle position according to the fitness function; Step S1.24: Update the extreme value of the particle individual position, that is, the historical optimal position of the particle individual: if the fitness of the current particle individual position is better than the global extreme value, that is, the optimal position of the current particle individual, then update the extreme value of the particle individual position ; Step S1.25: Update the global extreme value: If the fitness of the current particle individual position is better than the global extreme value, then update the global extreme value ; Step S1.26: Update individual particle speed: update the speed of each individual particle according to the individual particle position extreme value and the global extreme value; Step S1.27: updating the individual particle position according to the updated individual particle velocity; Step S1.28: Repeat the iteration steps S1.23 to S1.27 until the preset stop condition, i.e. the maximum number of iterations, is met; Step S1.29: After the preset stop condition is met, the individual particle position corresponding to the global extreme value is output as the optimal solution, i.e., the optimal feedforward model parameter; Step S2: inputting the feed water flow rate, the conductivity of the ammonia solution and the pH setting value at the economizer inlet into the improved particle swarm optimization feedforward model to generate a first predicted frequency of the ammonia pump; Step S3: inputting the difference between the actual pH value at the economizer inlet and the set pH value at the economizer inlet into the outer loop PI controller to generate the set pH value at the deaerator inlet, i.e., the set value of the inner loop PI controller; Step S4: inputting the difference between the actual pH value at the deaerator inlet and the set pH value at the deaerator inlet into the inner loop PI controller to generate a second predicted frequency of the ammonia addition pump; Step S5: Add the first predicted frequency of the ammonia adding pump and the second predicted frequency of the ammonia adding pump to obtain the total ammonia adding pump frequency, so as to realize automatic control of chemical ammonia addition.

[0005] Furthermore, the feedforward model is expressed as: ; In the formula, for The frequency of ammonia pump at the moment; for Water flow rate at the moment; for Economizer inlet pH set value at the moment; for Conductivity of ammonia solution at time.

[0006] Furthermore, in step S1.26, the update formula of the individual particle velocity is: ; ; In the formula, represents the inertia weight; and Respectively represent the maximum and minimum values ​​of the inertia weight; represents the base of natural logarithms; Indicates the current iteration number; Indicates the maximum number of iterations; represents the adjustment factor, which is 0.1; represents the Beta distribution; Both represent the shape parameters of the Beta distribution; Indicates The individual particles The speed of the moment; It is The individual particles The speed of the moment; and is the learning factor; and is a random number between [0,1]; express The individual position extreme value of the particle at the moment; express Global extremum at any moment.

[0007] Furthermore, in step S1.27, the individual particle position is updated according to the updated individual particle velocity, which is expressed as: ; ; In the formula, Indicates The individual particles The location at the moment; Represents the dynamic adjustment factor.

[0008] Furthermore, the specific process of step S3 is: the difference between the actual pH value at the economizer inlet and the pH set value at the economizer inlet is input into the outer loop PI controller, and the pH set value at the deaerator inlet is calculated by the proportional and integral controllers in the outer loop PI controller. , which is the setting value of the inner loop PI controller, is expressed as: ; In the formula, is the proportional gain coefficient of the outer loop PI controller; is the integral gain coefficient of the outer loop PI controller; for The difference between the set pH value at the economizer inlet and the actual pH value at the economizer inlet at the moment; for The difference between the set pH value at the economizer inlet and the actual pH value at the economizer inlet at a given moment.

[0009] Furthermore, the specific process of step S4 is: input the difference between the actual pH value of the deaerator inlet and the pH set value of the deaerator inlet into the inner loop PI controller, and calculate the second ammonia pump frequency by the proportional and integral controllers in the inner loop PI controller. , expressed as: ; In the formula, is the proportional gain coefficient of the inner loop PI controller; is the integral gain coefficient of the inner loop PI controller; express The difference between the pH set value at the deaerator inlet and the actual pH value at the deaerator inlet at the moment; for The difference between the deaerator inlet pH set value and the deaerator inlet pH actual value at the moment.

[0010] Furthermore, the outer loop PI controller calculates the deaerator inlet pH setpoint The inner loop PI controller calculates the frequency of the second ammonia pump When , a first-order inertia link is added for filtering.

[0011] An electronic device comprises a processor, a memory and a bus, wherein the processor and the memory are connected via the bus, wherein the memory is used to store a set of program codes, and the processor is used to call the program codes stored in the memory to execute a chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model.

[0012] A non-volatile computer storage medium stores computer executable instructions, which can execute a chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model.

[0013] Compared with the existing technology, the present invention has the following beneficial effects:

[0014] (1) The present invention realizes automatic control of water supply and ammonia addition by adopting a feedforward plus cascade PID algorithm, which greatly reduces manual workload and saves labor costs.

[0015] (2) The present invention determines the feedforward model structure by analyzing the time-lag relationship between the system mechanism and each variable, and uses an improved particle swarm algorithm to identify the feedforward model parameters, which can achieve accurate prediction of ammonia addition demand and enable timely and accurate ammonia addition.

[0016] (3) The present invention improves the inertia weight update formula by adopting a nonlinear attenuation method and a beta distribution function, constructs a dynamic adjustment factor by using the current individual extreme value and the global extreme value, adaptively adjusts the step size, improves the position update formula, and improves the global search capability of the particle swarm algorithm.

[0017] (4) The present invention is based on a cascade PID algorithm, which uses the pH value at the deaerator inlet as the inner loop control variable and the pH value at the economizer inlet as the outer loop control variable. It overcomes the nonlinearity and large time lag characteristics of the feed water ammonia addition system, can quickly respond to load changes, effectively suppress load interference, maintain the stability of the boiler feed water pH value, and ensure safe and economical operation of the unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 The figure is a flow chart of the method of the present invention.

[0019] Figure 2 The schematic diagram of the method of the present invention. DETAILED DESCRIPTION

[0020] like Figure 1-Figure 2 As shown, the present invention provides a technical solution: a chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model, comprising the following steps:

[0021] Step S1: Establish an improved particle swarm optimization feedforward model.

[0022] Step S2: Input the feed water flow rate, the conductivity of the ammonia solution and the set value of the pH at the economizer inlet into the improved particle swarm optimization feedforward model to generate the first predicted frequency of the ammonia dosing pump; the set value of the pH at the economizer inlet is usually a constant of 9.2.

[0023] Step S3: input the difference between the actual pH value at the economizer inlet and the set pH value at the economizer inlet into the outer loop PI controller to generate the set pH value at the deaerator inlet, ie, the set value of the inner loop PI controller.

[0024] Step S4: input the difference between the actual pH value at the deaerator inlet and the set pH value at the deaerator inlet into the inner loop PI controller to generate a second predicted frequency of the ammonia addition pump.

[0025] Step S5: Add the first predicted frequency of the ammonia adding pump and the second predicted frequency of the ammonia adding pump to obtain the total ammonia adding pump frequency, so as to realize automatic control of chemical ammonia addition.

[0026] The specific process of step S1 is as follows: by screening the variables related to the frequency of the ammonia pump during the operation of the thermal power unit, and considering the time lag relationship between the mechanism of the thermal power unit and each related variable, a feedforward model is constructed, and the improved particle swarm algorithm is used to identify the feedforward model parameters to obtain an improved particle swarm optimization feedforward model.

[0027] Among them, the feedforward model is expressed as: ; In the formula, for The frequency of ammonia pump at the moment; , , , , are the feedforward model parameters, respectively, the water flow lag time , conductivity lag time of ammonia solution , pH lag time at economizer inlet , weight of pH value at economizer inlet With bias ; for Water flow rate at the moment; for Economizer inlet pH set value at the moment; for Conductivity of ammonia solution at time.

[0028] Among them, the specific process of using the improved particle swarm algorithm to identify the parameters of the feedforward model is: Step S1.21: Based on the feedforward model parameters , , , , , set the improved particle swarm algorithm The position of each particle is The fitness function of the improved particle swarm algorithm is the prediction error of the frequency of the ammonia pump The variance of .

[0029] Step S1.22: Initialize the improved particle swarm algorithm; determine the number of individual particles, the initial position and speed of each individual particle.

[0030] Step S1.23: Calculate the fitness of each individual particle position according to the fitness function.

[0031] Step S1.24: Update the extreme value of the particle individual position (the historical optimal position of the particle individual): If the fitness of the current particle individual position is better than the global extreme value (the optimal position of the current particle individual), then update the extreme value of the particle individual position .

[0032] Step S1.25: Update the global extreme value: If the fitness of the current particle individual position is better than the global extreme value, then update the global extreme value .

[0033] Step S1.26: Update individual particle speed: Update the speed of each individual particle according to the individual particle position extreme value and the global extreme value. The update formula of the individual particle speed is: ; ; In the formula, represents the inertia weight; and Respectively represent the maximum and minimum values ​​of the inertia weight; represents the base of natural logarithms; Indicates the current iteration number; Indicates the maximum number of iterations; represents the adjustment factor, which is 0.1; represents the Beta distribution; Both represent the shape parameters of the Beta distribution. In this embodiment Take 1, Take 3; Indicates The individual particles The speed of the moment; It is The individual particles The speed of the moment; and is the learning factor; and is a random number between [0,1]; express The individual position extreme value of the particle at the moment; express Global extremum at any moment.

[0034] Step S1.27: Update the individual particle position according to the updated individual particle velocity, expressed as: ; ; In the formula, Indicates The individual particles The location at the moment; Represents the dynamic adjustment factor.

[0035] The present invention improves the inertia weight update formula by adopting a nonlinear attenuation method and a beta distribution function, constructs a dynamic adjustment factor by adopting the current individual extreme value and the global extreme value, adaptively adjusts the step size, improves the position update formula, and improves the global search capability of the particle swarm algorithm.

[0036] Step S1.28: Repeat the iteration steps S1.23 to S1.27 until the preset stop condition, i.e. the maximum number of iterations, is met.

[0037] Step S1.29: After the preset stop condition is met, the individual particle position corresponding to the global extreme value is output as the optimal solution, that is, the optimal feedforward model parameter.

[0038] In this embodiment, the optimal solution of the improved particle swarm optimization feedforward model parameters is shown in Table 1, and can be adjusted according to specific circumstances.

[0039] Table 1 Optimal solution of improved particle swarm optimization feedforward model parameters

[0040] The specific process of step S3 is as follows: the difference between the actual pH value at the economizer inlet and the set pH value at the economizer inlet is input into the outer loop PI controller, and the pH set value at the deaerator inlet is calculated by the proportional and integral controllers in the outer loop PI controller. , which is the setting value of the inner loop PI controller, is expressed as: ; In the formula, is the proportional gain coefficient of the outer loop PI controller; is the integral gain coefficient of the outer loop PI controller; for The difference between the set pH value at the economizer inlet and the actual pH value at the economizer inlet at the moment; for The difference between the set pH value at the economizer inlet and the actual pH value at the economizer inlet at a given moment.

[0041] The specific process of step S4 is as follows: the difference between the actual pH value at the deaerator inlet and the set pH value at the deaerator inlet is input into the inner loop PI controller, and the frequency of the second ammonia pump is calculated by the proportional and integral controllers in the inner loop PI controller. , expressed as: ; In the formula, is the proportional gain coefficient of the inner loop PI controller; is the integral gain coefficient of the inner loop PI controller; express The difference between the pH set value at the deaerator inlet and the actual pH value at the deaerator inlet at the moment; for The difference between the deaerator inlet pH set value and the deaerator inlet pH actual value at the moment.

[0042] In this embodiment, the parameters of the outer loop PI controller and the inner loop PI controller of the cascade PID algorithm are shown in Table 2 and can be adjusted according to actual conditions.

[0043] Table 2 Parameters of the outer loop PI controller and inner loop PI controller of the cascade PID algorithm

[0044] Among them, in the measurement link of the actual pH value at the economizer inlet and the actual pH value at the deaerator inlet, a first-order inertia link is added for filtering to suppress the random noise of the actual pH value at the economizer inlet and the actual pH value at the deaerator inlet. The first-order inertia link formula is It is expressed as: ; In the formula, is the Laplace operator; is the inertia time constant, in this embodiment Set to 2~5 to avoid Too low will result in poor filtering effect, or If the value is too high, it will cause excessive delay and can be adjusted according to the actual situation.

[0045] An electronic device comprises a processor, a memory and a bus, wherein the processor and the memory are connected via the bus, wherein the memory is used to store a set of program codes, and the processor is used to call the program codes stored in the memory to execute a chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model.

[0046] A non-volatile computer storage medium stores computer executable instructions, which can execute a chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model.

[0047] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A chemical ammonia addition cascade control method based on an improved particle swarm optimization feedforward model is characterized in that: The steps include: Step S1: Establishing an improved particle swarm optimization feedforward model; By screening the variables related to the frequency of the ammonia pump during the operation of the thermal power unit, and considering the time lag relationship between the mechanism of the thermal power unit and the relevant variables, a feedforward model is constructed, and the improved particle swarm algorithm is used to identify the parameters of the feedforward model, and the improved particle swarm optimization feedforward model is obtained; The specific process of using the improved particle swarm algorithm to identify the parameters of the feedforward model is as follows: Step S1.21: Based on the feedforward model parameters , , , , , set the improved particle swarm algorithm The position of each particle is The fitness function of the improved particle swarm algorithm is the prediction error of the frequency of the ammonia pump The variance of is the water flow lag time, is the conductivity lag time of ammonia solution, is the pH lag time at the economizer inlet, and are the weight and bias of the economizer inlet pH value, Indicates time; Step S1.22: Initialize the improved particle swarm algorithm; determine the number of individual particles, the initial position and speed of each individual particle; Step S1.23: Calculate the fitness of each individual particle position according to the fitness function; Step S1.24: Update the extreme value of the particle individual position, that is, the historical optimal position of the particle individual: if the fitness of the current particle individual position is better than the global extreme value, that is, the optimal position of the current particle individual, then update the extreme value of the particle individual position ; Step S1.25: Update the global extreme value: If the fitness of the current particle individual position is better than the global extreme value, then update the global extreme value ; Step S1.26: Update individual particle speed: update the speed of each individual particle according to the individual particle position extreme value and the global extreme value; Step S1.27: updating the individual particle position according to the updated individual particle velocity; Step S1.28: Repeat the iteration steps S1.23 to S1.27 until the preset stop condition, i.e. the maximum number of iterations, is met; Step S1.29: After the preset stop condition is met, the individual particle position corresponding to the global extreme value is output as the optimal solution, i.e., the optimal feedforward model parameter; Step S2: inputting the feed water flow rate, the conductivity of the ammonia solution and the pH setting value at the economizer inlet into the improved particle swarm optimization feedforward model to generate a first predicted frequency of the ammonia pump; Step S3: inputting the difference between the actual pH value at the economizer inlet and the set pH value at the economizer inlet into the outer loop PI controller to generate the set pH value at the deaerator inlet, i.e., the set value of the inner loop PI controller; Step S4: inputting the difference between the actual pH value at the deaerator inlet and the set pH value at the deaerator inlet into the inner loop PI controller to generate a second predicted frequency of the ammonia addition pump; Step S5: Add the first predicted frequency of the ammonia adding pump and the second predicted frequency of the ammonia adding pump to obtain the total ammonia adding pump frequency, so as to realize automatic control of chemical ammonia addition.

2. The chemical ammonia addition cascade control method based on the improved particle swarm optimization feedforward model according to claim 1 is characterized in that: The feedforward model is expressed as: ； In the formula, for The frequency of ammonia pump at the moment; for Water flow rate at the moment; for Economizer inlet pH set value at the moment; for Conductivity of ammonia solution at time.

3. The chemical ammonia addition cascade control method based on the improved particle swarm optimization feedforward model according to claim 2 is characterized in that: In step S1.26, the update formula of the particle individual velocity is: ； ； In the formula, represents the inertia weight; and Respectively represent the maximum and minimum values ​​of the inertia weight; represents the base of natural logarithms; Indicates the current iteration number; Indicates the maximum number of iterations; represents the adjustment factor, which is 0.1; represents the Beta distribution; Both represent the shape parameters of the Beta distribution; Indicates The individual particles The speed of the moment; It is The individual particles The speed of the moment; and is the learning factor; and is a random number between [0,1]; express The individual position extreme value of the particle at the moment; express Global extremum at any moment.

4. The chemical ammonia addition cascade control method based on the improved particle swarm optimization feedforward model according to claim 3 is characterized in that: In step S1.27, the individual particle position is updated according to the updated individual particle velocity, which is expressed as: ； ； In the formula, Indicates The individual particles The location at the moment; Represents the dynamic adjustment factor.

5. The chemical ammonia addition cascade control method based on the improved particle swarm optimization feedforward model according to claim 4 is characterized in that: The specific process of step S3 is: input the difference between the actual pH value at the economizer inlet and the set pH value at the economizer inlet into the outer loop PI controller, and calculate the set pH value at the deaerator inlet through the proportional and integral controllers in the outer loop PI controller. , which is the setting value of the inner loop PI controller, is expressed as: ； In the formula, is the proportional gain coefficient of the outer loop PI controller; is the integral gain coefficient of the outer loop PI controller; for The difference between the set pH value at the economizer inlet and the actual pH value at the economizer inlet at the moment; for The difference between the set pH value at the economizer inlet and the actual pH value at the economizer inlet at a given moment.

6. The chemical ammonia addition cascade control method based on the improved particle swarm optimization feedforward model according to claim 5 is characterized in that: The specific process of step S4 is: input the difference between the actual pH value at the deaerator inlet and the set pH value at the deaerator inlet into the inner loop PI controller, and calculate the frequency of the second ammonia pump through the proportional and integral controllers in the inner loop PI controller. , expressed as: ； In the formula, is the proportional gain coefficient of the inner loop PI controller; is the integral gain coefficient of the inner loop PI controller; express The difference between the pH set value at the deaerator inlet and the actual pH value at the deaerator inlet at the moment; for The difference between the deaerator inlet pH set value and the deaerator inlet pH actual value at the moment.

7. The chemical ammonia addition cascade control method based on the improved particle swarm optimization feedforward model according to claim 6 is characterized in that: Calculate the deaerator inlet pH setpoint in the outer loop PI controller The inner loop PI controller calculates the frequency of the second ammonia pump When , a first-order inertia link is added for filtering.

8. An electronic device, characterized in that: The method comprises a processor, a memory and a bus, wherein the processor and the memory are connected via the bus, wherein the memory is used to store a set of program codes, and the processor is used to call the program codes stored in the memory to execute the chemical ammonia addition cascade control method based on the improved particle swarm optimization feedforward model as described in any one of claims 1 to 7.

9. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions execute the chemical ammonia addition cascade control method based on the improved particle swarm optimization feedforward model described in any one of claims 1-7.

Images