CN112882387A - Control strategy design method of bag type dust collector - Google Patents
Control strategy design method of bag type dust collector Download PDFInfo
- Publication number
- CN112882387A CN112882387A CN202110038227.5A CN202110038227A CN112882387A CN 112882387 A CN112882387 A CN 112882387A CN 202110038227 A CN202110038227 A CN 202110038227A CN 112882387 A CN112882387 A CN 112882387A
- Authority
- CN
- China
- Prior art keywords
- value
- frequency
- state
- sliding mode
- state value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
- Y02B30/00—Energy efficient heating, ventilation or air conditioning [HVAC]
- Y02B30/70—Efficient control or regulation technologies, e.g. for control of refrigerant flow, motor or heating
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a control strategy design method of a bag type dust collector, which constructs a stable mathematical model of frequency converter frequency, electric control valve opening degree, airflow flow and air pressure difference of a gas purification chamber on the basis of two data acquisition schemes of a stable state and a transient state of the bag type dust collector, acquires open-loop control experimental process data through a transient state process, and provides a model and data support for a subsequent control design scheme. The design method adopts an immune algorithm to identify the parameter matrix in the state space equation, so that the overall convergence of the sliding mode control strategy is improved; the discrete sliding mode variable structure control method based on the approach law is adopted, the states of the controlled airflow flow and the air pressure difference of the air purifying chamber in the movement of the switching surface are effectively analyzed, and the whole bag type dust collector system is enabled to keep good stability in the state space.
Description
Technical Field
The invention relates to the field of bag type dust collectors, in particular to a control strategy design method of a bag type dust collector.
Background
The bag type dust collector is widely applied by virtue of high-efficiency dust collection capacity. The bag type dust collector mainly comprises two processes of dust filtration and dust removal during working, an electric control valve at an air outlet needs to be started at first in the dust filtration process, an induced draft fan starts to operate to enable dust-containing airflow to penetrate through a filter bag, and dust can be filtered by the filter bag. The ash removal means that pulse compressed air is sprayed through a nozzle to remove dust accumulated on a filter bag, so that the filter bag recovers the dust removal and filtration capacity again. In order to achieve a high dust removal efficiency of the bag house, the system operation state of the bag house needs to be controlled so that the output variable thereof becomes a target state value. Most of control systems of the bag type dust collector adopt open-loop control, the control system is a system control mode without feedback information, the traditional open-loop control method easily causes the problems of long adjusting time, large steady-state error and the like, the energy consumption is increased, the operation efficiency is reduced, and the stable operation of the dust collection process is not facilitated.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention aims to provide a control strategy design method of a bag type dust collector. The design method of the invention combines the internal flow field motion principle of the bag type dust collector and the characteristics of the control system to select the airflow flow and the air pressure difference of the air purifying chamber for research. The frequency of a frequency converter and the opening of an electric control valve in the bag type dust collector are used as input, and airflow and air pressure difference of a clean air chamber are used as output. The overall control process of the bag-type dust collector is analyzed from two aspects of a steady state and a transient state, and in a steady-state data acquisition experimental scheme, the aim is to determine the numerical values after the flow and the pressure difference reach stable values when the frequency of a frequency converter and the opening of an electric control valve are a certain set of constant values; the process is to combine the selectable numerical value of the frequency converter and the selectable numerical value of the opening of the electric control valve, set corresponding numerical values on a test platform, and record the steady-state numerical values when the flow and the pressure difference reach the steady state; and according to the result of the steady-state data acquisition experiment scheme, establishing a steady-state mathematical model of the frequency converter, the opening of the electric control valve, the airflow and the air pressure difference of the air purifying chamber. In the transient data acquisition experimental scheme, the frequency of the frequency converter and the opening of the electric control valve are directionally adjusted according to the model, and the change generated by the airflow flow and the air pressure difference of the air purifying chamber in the whole change process from the initial state to the stable state of the control system is recorded. The transient control process of the bag type dust collector is further analyzed on the basis of the steps, a state space equation is used as a system model, parameter identification is carried out on the state space model through an immune algorithm, a parameter matrix of a discrete system is further obtained, the control method selects discrete sliding mode variable structure control based on an approximation law, the difference value of the expected state value of the bag type dust collector system and the discrete sliding mode switching surface is calculated, a discrete index approximation law and a discrete sliding mode switching surface are designed for the discrete control system of the bag type dust collector, a discrete sliding mode control law is further designed, and the control law is applied to the discrete sliding mode variable structure control, so that the sliding mode controller of the bag type dust collector is obtained. The sliding mode controller of the bag type dust collector obtained by the design method enhances the overall control effect of the sliding mode controller, effectively reduces buffeting of a bag type dust collector system in the regulation and control process, reduces regulation and control errors and shortens the regulation time.
The technical scheme for solving the technical problems is to design a control strategy design method of a bag type dust collector, which is characterized by comprising the following steps of:
the method comprises the following steps: selecting the frequency of a frequency converter of an induced draft fan arranged at the outlet of an air outlet pipeline and the opening of an electric control valve arranged at the middle position in the air outlet pipeline as input variables of a control strategy by combining the dust removal principle of a bag type dust remover, taking the airflow flow of the air outlet pipeline at the inner side of the electric control valve and the air pressure difference of a clean air chamber as output variables, testing the output variables corresponding to different input variables by adjusting the input variables to obtain a stable state mathematical model of the output variables, and taking the value of the output variable obtained by the minimum input variable as the initial state value of the system;
step two: obtaining experimental data of transient process
Setting a target state value of a dust remover system, substituting the target state value into the steady-state mathematical model of the output variable established in the step one, and calculating to obtain a value of an input variable corresponding to the target state value, namely obtaining the frequency of a target frequency converter and the opening of a target electric control valve; then opening a bag type dust collector system, adjusting the opening of the electric control valve to a target opening of the electric control valve, adjusting the frequency of the frequency converter to a value which is 5-10 Hz lower than the frequency of the target frequency converter, keeping the opening of the electric control valve unchanged, directly adjusting the output frequency of the frequency converter from a current value to a target frequency value of the frequency converter, and simultaneously recording the airflow flow and the air pressure of the air purification chamber at different moments in the processObtaining experimental data of the transient process, wherein the experimental data of the transient process is the state value of the transient process system; the sampling interval of the experimental data is 0.05s, until the numerical values of the airflow flow and the air pressure difference of the air purifying chamber reach the target state value, the state values of the system at K moments are obtained, and for any moment K, the state value of the system is as follows: x (k) ═ x1(k),x2(k)]T，x1(k) Representing the flow rate of the gas stream, x2(k) Representing the air pressure difference of the air purifying chamber;
step three: designing a discrete sliding mode control strategy
1) Determining a discrete-time system state space equation:
x(k+1)＝Ax(k)+bu(k) (3.1)
in the above formula, the system state variable is x (k) ═ x1(k),x2(k)]T，x1(k) Representing the flow rate of the gas stream, x2(k) Representing the net chamber differential pressure, the system input variable is u (k) ═ u1(k),u2(k)]T，u1(k) Representing the frequency of the frequency converter, u2(k) Indicating the opening of the electric control valve; k corresponds to any sampling time point;
2) identifying the parameter A, b in the state space equation by adopting an immune algorithm, selecting the number of immune individuals as 80, the maximum immune algebra G as 400 and the variation probability pm as 0.7, wherein the specific process comprises the following steps:
2.1 setting the value range of the parameter A, b, generating an initial antibody by a random method, determining the initial antibody as the initial value of the parameter A, b, and setting the target function as the difference value between the predicted state value and the actual state value after substituting the antibody into the state space equation of the discrete time system;
2.2 Euclidean distance between antibodies was used to calculate affinity:
wherein abi,kDenotes the k dimension, ab of antibody ij,kDenotes the kth dimension of antibody j, L denotes the total antibody-encoding dimension, L ═ 6;
2.3 antibody concentration calculation algorithm:
wherein N is the number of immunized individuals and is 80; s (ab)i,abj) Is the similarity of the antibodies:
wherein δ is 0.2;
2.4 calculate encouragement values from affinity and antibody concentration:
sorting the encouragement values from high to low, and selecting the first 50% of the sorted individuals for reservation;
2.5 continuously generating antibodies of new generation for the reserved individuals through the actions of cloning, mutation and clone inhibition operators, reserving the first 50% of the individuals with the encouragement value from high to low in the antibodies generated in each generation, repeating the actions of cloning, mutation and clone inhibition operators to generate antibodies of the next generation until the immunization generation G reaches 400, reserving the individuals with the highest encouragement value obtained in the last generation to obtain each parameter matrix;
3) establishing a discrete sliding mode control law
The control method selects a sliding mode variable structure control method based on a discrete index approach law, and the selected switching surface is as follows:
s(k)＝E-Cx(k) (3.6)
wherein the coefficient matrix of the switching plane is C ═ 5, 1; e represents a target state value, and defines a switching zone including a switching surface on the basis of the switching surface s (k):
SΔ＝{x∈Rn|-Δ＜s(k)＜+Δ} (3.7)
2 Δ is the width of the switching band, Δ is 0.01; when determining the control law u (k) in the state space equation of the discrete variable structure system, firstly defining the form of discrete exponential approximation law:
s(k+1)-s(k)＝-qTs(k)-εTsat(s(k)),ε＞0,q＞0 (3.8)
wherein T is the sampling period;
according to the equation of state space, the switching surface and the discrete index approach law, the discrete sliding mode control law is obtained as follows:
u(k)＝-(Cb)-1[CAx(k)-(1-qT)s(k)+εTsat(s)] (3.10)
in the above formula, sat(s) is a saturation function, which is used to replace the sign function in the original approach law, and the function is:
step four: the method comprises the following steps of (1) searching the optimal values of the coefficients q and epsilon of the discrete sliding mode control law by using a genetic algorithm to obtain the discrete sliding mode control law, wherein the specific process comprises the following steps:
4.1 substituting the parameters A and b of the state space equation obtained in the step three into a discrete sliding mode control law; setting the value ranges of the discrete sliding mode control law coefficients q and epsilon;
4.2, coding two unknown discrete sliding mode control law coefficients q and epsilon, and converting decimal numbers into binary character strings, wherein the length of the binary character strings is 9;
4.3 importing the state value x (1) of the experimental data of the transient process at the first moment acquired in the step two, substituting x (1) and a group of initial q and epsilon into a discrete sliding mode control law to obtain u (1), substituting u (1) into a state space equation, and calculating the system state x at the second moment of simulationpro(2) By calculating sequentially, the simulated state value x at all times is obtainedpro(3),xpro(4),...,xpro(K) Wherein x ispro(k) Is the analog state value at the moment of K, K is more than or equal to K>1; x to be simulatedpro(k) The difference between the target state value E and the target state value E is recorded as r, and the target function o of the discrete sliding mode control law coefficients q and epsilonbjvsc is:
objvsc＝r＝∑[E-xpro(k)] (4.1)
the control purpose of the dust remover system is to reduce the value of the objective function objvsc, and to set the fitness function Fit to be the same as the objective function objvsc; the value of the objective function objvsc calculated by each group q and epsilon is the value of the fitness function Fit corresponding to a group q and epsilon;
4.4 setting the initial individual number to be 40, and adopting a random traversal sampling algorithm by a selection operator in the genetic algorithm:
where Fit (x)i) For each individual fitness value, F (x)i) Is an individual xiThe selected probability value N is the total number of the individuals;
4.5 selecting a single-point crossing algorithm, wherein the crossing probability is 0.6; selecting an algorithm of basic bit variation, wherein the variation probability is 0.05; carrying out cross variation on the initial 40 individuals to obtain a new generation of individuals, setting the number of genetic iteration individuals to be 50, selecting 50 individuals with higher selected probability to carry out cross variation to generate a next generation of individuals once in each genetic iteration, and continuously iterating until the number of the genetic iteration reaches 100; outputting the individuals with the highest selected probability in the individuals generated by the last genetic iteration, decoding the individuals according to the step 4.2, and calculating to obtain values of q and epsilon;
4.6 substituting the values of q and epsilon obtained in the step 4.5 into a discrete sliding mode control law to obtain the discrete sliding mode control law without unknown coefficients;
step five: starting the bag type dust collector system, and adjusting the frequency of a frequency converter of the bag type dust collector system and the opening of an electric control valve to be the value of an input variable corresponding to the initial state value of the system; according to a set target state value E of the bag type dust collector system, a switching surface and a discrete sliding mode control law without unknown coefficients, the frequency and the opening degree of an electric control valve of a frequency converter to be achieved by first regulation are calculated by combining with a system initial state value, the system state value is monitored, the system state value is collected once at intervals of a certain time, the frequency and the opening degree of the electric control valve of the frequency converter to be achieved by next regulation are calculated according to the system state value, the frequency and the opening degree of the electric control valve of the frequency converter are continuously monitored and calculated, the frequency and the opening degree of the electric control valve of the frequency converter are continuously regulated and controlled until the system state value reaches the target state value E, and accordingly, the control of the bag type dust collector system is completed.
Compared with the prior art, the invention has the beneficial effects that:
1. the design method of the invention constructs a stable mathematical model of the output frequency of the frequency converter, the opening of the electric control valve, the airflow and the air pressure difference of the air purifying chamber on the basis of two data acquisition schemes of the stable state and the transient state of the bag type dust collector, acquires the data of the open-loop control experiment process through the transient state process, and provides a model and a data support for the subsequent control design scheme.
2. The design method adopts an immune algorithm to identify the parameter matrix in the state space equation, so that the overall convergence of the sliding mode control strategy is improved; the discrete sliding mode variable structure control method based on the approach law is adopted, the states of the controlled airflow flow and the air pressure difference of the air purifying chamber in the movement of the switching surface are effectively analyzed, and the whole bag type dust collector system is enabled to keep good stability in the state space.
3. The design method improves the sliding mode variable structure control method, optimizes the discrete index approach law parameters by using a genetic algorithm, adds a saturation function to replace a sign function in the original approach law, enhances the overall control effect of the sliding mode control strategy, effectively reduces buffeting of a bag type dust collector system in the regulation and control process, reduces regulation and control errors and shortens the regulation time.
Drawings
FIG. 1 is a flow chart of an embodiment of a design method of the present invention.
FIG. 2 is a schematic diagram of a sliding mode variable structure control method of a discrete index approach law.
Fig. 3 is a schematic plan view of an experimental device simulating the working principle of a bag-type dust collector (only used for illustrating the dust-collecting principle of the bag-type dust collector).
Detailed Description
The invention will be further described with reference to the following figures and specific examples, which are not intended to limit the invention in any way.
Referring to fig. 3, an induced draft fan 4 is installed at an air outlet and has a function of maintaining the air flow of the whole dust collector system in the filtering operation process, a frequency converter is used for adjusting the rotating speed of the induced draft fan, and when the output frequency of the frequency converter is changed, the rotating speed of the induced draft fan 4 is changed accordingly, so that the purposes of adjusting the flow in the dust collector system and adjusting the air pressure difference of a clean air chamber are achieved. The electric control valve 3 is arranged in the pipeline at the air outlet, and the opening of the valve can be adjusted, so that the flow in the dust remover system and the air pressure difference of the air purification chamber can be changed. The output variable data is obtained by a data collecting system, the data collecting system is composed of a sensor, a signal collecting module and an industrial personal computer, the sensor for collecting the airflow is a vortex flow meter 2, and the vortex flow meter 2 is installed in an air outlet pipeline on the inner side of the electric control valve 3 to obtain the airflow detection data of the air outlet pipeline. The 2 model of the vortex shedding flowmeter is ZY-LU, and the measuring range is 50m3/h-3500m3The drift diameter is 300 mm. The air pressure difference of the air purifying chamber refers to the difference between the air pressure in the air purifying chamber and the external atmospheric pressure, the sensor for collecting the air pressure difference of the air purifying chamber is a micro-pressure transmitter 1, and the micro-pressure transmitter 1 is installed on the inner side and the outer side of the air purifying chamber to obtain the detection data of the air pressure difference of the air purifying chamber. The micro-pressure transmitter 1 is AST47LP, and the measuring range is-15 kpa to 1 Mpa. The sampling interval of the data of the vortex shedding flowmeter 2 and the micro-pressure transmitter 1 is 0.05s, the vortex shedding flowmeter and the micro-pressure transmitter are both connected with an information acquisition module, and the information acquisition module converts the data acquired by the two sensors and then sends the converted data to the memory of the industrial personal computer.
The invention provides a design method of a control strategy of a bag type dust collector (the design method is called as a design method for short, and see the figures 1-2), which is characterized by comprising the following steps:
the method comprises the following steps: the method is characterized in that the output frequency of a frequency converter of an induced draft fan arranged at the outlet of an air outlet pipeline and the opening of an electric control valve arranged at the middle position in the air outlet pipeline are selected as input variables of a control strategy by combining the dust removal principle of a bag type dust remover, the airflow flow and the air pressure difference of an air outlet pipeline at the inner side (taking the outlet of the air outlet pipeline as the outer side) of the electric control valve are used as output variables, the output variables corresponding to different input variables are tested by adjusting the input variables to obtain a stable mathematical model of the output variables, and the value of the output variable obtained by the minimum input variable is used as the initial state value of the system.
The specific process of obtaining the steady-state mathematical model of the output variable is as follows:
1) selecting 16 induced draft fan frequency converters with the value range of 15 Hz-60 Hz, and selecting 12 electric control valves with the opening range of 5% -60%; expressing the frequency of the frequency converter as XiThe opening of the electrically controlled valve is represented as Yj. The specific combination form of the selected frequency of the frequency converter and the opening of the electric control valve is as follows:
sequentially combining the frequency of each frequency converter and the numerical value of the opening of the electric control valve to obtain 192 combined input conditions in total;
2) initializing a parameter variable of a dust remover system; and starting the dust remover system, adjusting the frequency of the frequency converter to 15Hz, adjusting the opening of the electric control valve to 5%, monitoring the value of the output variable of the dust remover system at the moment, and when the value of the output variable is stable and does not change, determining the value of the output variable at the moment as the initial state value of the system. The airflow rate monitored by the embodiment to obtain the initial state value of the system is 425m3·h-1The air pressure difference of the air purifying chamber is-342 Pa;
3) sequentially adjusting input variables of the dust collector system according to the 192 combination input conditions obtained in the step 1), recording the value of the output variable measured corresponding to each combination input condition (for preventing error data interference, each combination input condition is subjected to a plurality of experimental tests, recording is carried out after the output variable is stable, the average value of the output variables of the plurality of tests is taken as the value of the output variable corresponding to the input condition), analyzing and fitting the data of the two input variables and the two output variables (the data can be subjected to Python analysis and fitting), and constructing a steady-state mathematical model.
The steady state mathematical model of the airflow rate is:
the steady state mathematical model of the net air chamber pressure difference is as follows:
in the above two steady-state mathematical models, xbRepresenting the frequency of the frequency converter, xdIndicating the opening of the electrically controlled valve, F (x)b,xd) Represents the flow rate of the air flow in m3·h-1，P(xb,xd) Representing the net air chamber pressure difference in Pa.
Step two: obtaining experimental data of transient process
Set the target state value for the precipitator system (in this example, the airflow rate is 535 m)3·h-1And the air pressure difference of the air purifying chamber is-478 Pa), substituting the target state value into the steady-state mathematical model of the output variable established in the step one, and calculating to obtain the value of the input variable corresponding to the target state value, namely the frequency of the target frequency converter and the opening of the target electric control valve (in the embodiment, the frequency of the target frequency converter corresponding to the target state value of the system is 30Hz, and the opening of the target electric control valve is 40%). And then opening the bag type dust collector system, adjusting the opening of the electric control valve to a target opening of the electric control valve, adjusting the frequency of the frequency converter to a value (25 Hz in the embodiment) which is 5-10 Hz smaller than the frequency of the target frequency converter, keeping the opening of the electric control valve unchanged, directly adjusting the frequency of the frequency converter from the current value to the frequency value of the target frequency converter, and simultaneously recording the airflow and the air pressure difference of the air purification chamber at different moments in the process to obtain experimental data of the transient process, wherein the experimental data of the transient process is the state value of the transient process system. The sampling interval of the experimental data was 0.05s until the flow rate and the net flowThe value of the air pressure difference of the air chamber reaches a target state value, the system state values at K moments are obtained, and for any moment K, the system state value is as follows: x (k) ═ x1(k),x2(k)]T，x1(k) Representing the flow rate of the gas stream, x2(k) Representing the net air chamber pressure differential.
Step three: designing a discrete sliding mode control strategy
1) Determining a discrete-time system state space equation:
x(k+1)＝Ax(k)+bu(k) (3.1)
in the above formula, the system state variable is x (k) ═ x1(k),x2(k)]T，x1(k) Representing the flow rate of the gas stream, x2(k) Representing the net chamber differential pressure, the system input variable is u (k) ═ u1(k),u2(k)]T，u1(k) Representing the frequency of the frequency converter, u2(k) Indicating the electronically controlled valve opening. k corresponds to any sampling time point in the process of acquiring experimental data of the transient process.
2) Identifying the parameter A, b in the state space equation by adopting an immune algorithm, selecting the number of immune individuals as 80, the maximum immune algebra G as 400 and the variation probability pm as 0.7, wherein the specific process comprises the following steps:
2.1 setting the value range of the parameter A, b, wherein the element value range is [ -2,2], generating an initial antibody by a random method, determining the initial antibody as the initial value of the parameter A, b, and determining the objective function as the difference between the predicted state value after substituting the antibody into the discrete time system state space equation and the actual state value (the actual state value is the experimental data of the transient process obtained in the second step, and the value of u (k) at each time is the value of the input variable corresponding to the target state value during the prediction calculation in the second step).
2.2 Euclidean distance between antibodies was used to calculate affinity:
wherein abi,kDenotes the k dimension, ab of antibody ij,kDenotes the kth dimension of antibody j, L denotes the total antibody-encoding dimension, and L ═ 6.
2.3 antibody concentration calculation algorithm:
wherein N is the number of immunized individuals and is 80. S (ab)i,abj) Is the similarity of the antibodies:
wherein δ is 0.2.
2.4 calculate encouragement values from affinity and antibody concentration:
the encouragement values are ranked from high to low and the top 50% of the ranked individuals are selected for retention.
2.5 continuously generating antibodies of new generation for the reserved individuals through the actions of cloning, mutation and clone inhibition operators, reserving the first 50% of the individuals with the encouragement value from high to low in the antibodies generated in each generation, repeating the actions of cloning, mutation and clone inhibition operators to generate antibodies of the next generation until the immunization generation G reaches 400, reserving the individuals with the highest encouragement value obtained in the last generation, and obtaining each parameter matrix as follows:
3) establishing a discrete sliding mode control law
The control method selects a sliding mode variable structure control method based on a discrete index approach law, and the selected switching surface is as follows:
s(k)＝E-Cx(k) (3.6)
in which the system of switching surfacesThe number matrix is C ═ 5,1]. E represents a target state value, and in this embodiment, E is [535, -478 ]]TOn the basis of the switching surface s (k), a switching zone including the switching surface is defined:
SΔ＝{x∈Rn|-Δ＜s(k)＜+Δ} (3.7)
2 Δ is the width of the switching band, and Δ is 0.01. The movement of the system first approaches the switching zone from the initial state and then reaches the switching surface s (k) in limited steps and moves thereon. When determining the control law u (k) in the state space equation of the discrete variable structure system, firstly defining the form of discrete exponential approximation law:
s(k+1)-s(k)＝-qTs(k)-εTsat(s(k)),ε＞0,q＞0 (3.8)
wherein T is a sampling period, which is 0.05s in this embodiment, and the ideal quasi-sliding mode satisfies:
s(k+1)＝s(k)＝0,k＝0,1,2... (3.9)
according to the equation of state space, the switching surface and the discrete index approach law, the discrete sliding mode control law is obtained as follows:
u(k)＝-(Cb)-1[CAx(k)-(1-qT)s(k)+εTsat(s)] (3.10)
in the above formula, sat(s) is a saturation function, which is used to replace the sign function in the original approach law, and the function is:
step four: the method comprises the following steps of (1) searching the optimal values of the coefficients q and epsilon of the discrete sliding mode control law by using a genetic algorithm to obtain the discrete sliding mode control law, wherein the specific process comprises the following steps:
and 4.1 substituting the parameters A and b of the state space equation obtained in the step three into a discrete sliding mode control law. And setting the value ranges of the discrete sliding mode control law coefficients q and epsilon, wherein the value range of q is set to be [8,12], and the value range of epsilon is set to be [0,1 ].
4.2, coding two unknown discrete sliding mode control law coefficients q and epsilon, and converting decimal numbers into binary character strings, wherein the length of the binary character strings is 9.
4.3 importing the state value x (1) of the experimental data of the transient process at the first moment acquired in the step two, substituting the x (1) and a group of initial q and epsilon (the initial q and epsilon are randomly and repeatedly valued within the value range in the step 4.1) into a discrete sliding mode control law to obtain u (1), substituting the u (1) into a state space equation, and calculating the system state x at the second moment of simulationpro(2) By calculating sequentially, the simulated state value x at all times is obtainedpro(3),xpro(4),...,xpro(K) Wherein x ispro(k) Is the analog state value at the moment of K, K is more than or equal to K>1. X to be simulatedpro(k) And the target state value E (in the present embodiment, the flow rate of the air is 535 m)3·h-1The air pressure difference of the air purifying chamber is-478 Pa) is recorded as r, and the objective functions objvsc of the discrete sliding mode control law coefficients q and epsilon are as follows:
objvsc＝r＝∑[E-xpro(k)] (4.1)
the control purpose of the dust remover system is to reduce the value of the objective function objvsc, and to set the fitness function Fit to be the same as the objective function objvsc. The value of the objective function objvsc calculated by each group q and epsilon is the value of the fitness function Fit corresponding to each group q and epsilon.
4.4 set the initial number of individuals to 40 (i.e. 40 sets of initial q and epsilon which are not randomly identical), the selection operator in the genetic algorithm adopts a random traversal sampling algorithm:
where Fit (x)i) For each individual fitness value, F (x)i) Is an individual xiAnd the selected probability value N is the total number of the individuals.
4.5 selecting a single-point crossing algorithm, wherein the crossing probability is 0.6; selecting an algorithm of basic bit variation, wherein the variation probability is 0.05; and (3) carrying out cross variation on the initial 40 individuals to obtain a new generation of individuals, setting the number of genetic iteration individuals to be 50, carrying out cross variation on 50 individuals with high selected probability once in each genetic iteration to generate a next generation of individuals, and continuously iterating until the number of the genetic iteration reaches 100. And outputting the individuals with the highest selected probability in the individuals generated by the last genetic iteration, decoding the individuals according to the step 4.2, and calculating to obtain the values of q and epsilon. This example calculates q to be 10 and ε to be 0.35.
And 4.6, substituting the values of q and epsilon obtained in the step 4.5 into a discrete sliding mode control law to obtain the discrete sliding mode control law without unknown coefficients.
Step five: and starting the bag type dust collector system, and adjusting the frequency of a frequency converter of the bag type dust collector system and the opening of the electric control valve to be the value of the input variable corresponding to the initial state value of the system. According to a set target state value E of the bag type dust collector system, a switching surface (namely formula (3.6)) and a discrete sliding mode control law (namely formula (3.10)) without unknown coefficients, calculating the frequency of the frequency converter and the opening of the electric control valve which are required to be achieved by the first regulation and control by combining with a system initial state value, monitoring the system state value, collecting the system state value once at intervals of a certain time (set to be 0.5s), calculating the frequency of the frequency converter and the opening of the electric control valve which are required to be achieved by the next regulation and control according to the system state value, continuously monitoring, continuously calculating, continuously regulating and controlling the frequency of the frequency converter and the opening of the electric control valve until the system state value reaches the target state value E, and accordingly completing the control of the bag type dust collector system.
As a comparative example of the discrete sliding mode control strategy obtained by the design method, the system state values of different moments after the input variable value corresponding to the initial state value of the system of the bag-type dust collector is directly adjusted to the frequency of the target frequency converter and the opening degree of the target electric control valve, the time required for the system state value to reach the target state value E, the analysis time and the stability of the system state value in the process are recorded. The result shows that the time for regulating and controlling the bag-type dust collector system to the target state value is shorter by adopting the discrete sliding mode control strategy obtained by the design method of the invention, and the change of the system state value in the regulating and controlling process is more stable. Specifically, in terms of flow control, the time taken for the comparative example was 12.5s, and the steady state error was 5.9%; the time used by the discrete sliding mode control method is 9.5s, and the steady-state error is 3.5%; in the aspect of air purifying chamber pressure difference control, the time of the comparative example is 15.5s, and the steady-state error is 6.8%; the time used by the discrete sliding mode control method is 12.4s, and the steady-state error is 3.3%.
The sliding mode control strategy obtained by the design method can regulate and control the bag-type dust collector of the same model at any system state value until the target state value is reached; for different bag type dust collectors, a corresponding sliding mode control strategy can be obtained by referring to the design method. The design method lays a foundation for efficient and stable control of the state of the bag type dust collector, effectively reduces system buffeting, improves control efficiency, and well improves adjusting time and target state steady-state errors.
Nothing in this specification is said to apply to the prior art.
Claims (5)
1. A control strategy design method of a bag type dust collector is characterized by comprising the following steps:
the method comprises the following steps: selecting the frequency of a frequency converter of an induced draft fan arranged at the outlet of an air outlet pipeline and the opening of an electric control valve arranged at the middle position in the air outlet pipeline as input variables of a control strategy by combining the dust removal principle of a bag type dust remover, taking the airflow flow of the air outlet pipeline at the inner side of the electric control valve and the air pressure difference of a clean air chamber as output variables, testing the output variables corresponding to different input variables by adjusting the input variables to obtain a stable state mathematical model of the output variables, and taking the value of the output variable obtained by the minimum input variable as the initial state value of the system;
step two: obtaining experimental data of transient process
Setting a target state value of a dust remover system, substituting the target state value into the steady-state mathematical model of the output variable established in the step one, and calculating to obtain a value of an input variable corresponding to the target state value, namely obtaining the frequency of a target frequency converter and the opening of a target electric control valve; then opening the bag type dust collector system, adjusting the opening of the electric control valve to the target opening of the electric control valve, and adjusting the frequency of the frequency converter to be smaller than the target frequency conversionThe frequency of the frequency converter is 5-10 Hz, the opening of the electric control valve is kept unchanged, the frequency of the frequency converter is directly adjusted to a target frequency value of the frequency converter from a current value, and meanwhile, the airflow and the air pressure difference of the air purification chamber at different moments in the process are recorded to obtain experimental data of a transient process, wherein the experimental data of the transient process is a state value of a transient process system; the sampling interval of the experimental data is 0.05s, until the numerical values of the airflow flow and the air pressure difference of the air purifying chamber reach the target state value, the state values of the system at K moments are obtained, and for any moment K, the state value of the system is as follows: x (k) ═ x1(k),x2(k)]T，x1(k) Representing the flow rate of the gas stream, x2(k) Representing the air pressure difference of the air purifying chamber;
step three: designing a discrete sliding mode control strategy
1) Determining a discrete-time system state space equation:
x(k+1)＝Ax(k)+bu(k) (3.1)
in the above formula, the system state variable is x (k) ═ x1(k),x2(k)]T，x1(k) Representing the flow rate of the gas stream, x2(k) Representing the net chamber differential pressure, the system input variable is u (k) ═ u1(k),u2(k)]T，u1(k) Representing the frequency of the frequency converter, u2(k) Indicating the opening of the electric control valve; k corresponds to any sampling time point;
2) identifying the parameter A, b in the state space equation by adopting an immune algorithm, selecting the number of immune individuals as 80, the maximum immune algebra G as 400 and the variation probability pm as 0.7, wherein the specific process comprises the following steps:
2.1 setting the value range of the parameter A, b, generating an initial antibody by a random method, determining the initial antibody as the initial value of the parameter A, b, and setting the target function as the difference value between the predicted state value and the actual state value after substituting the antibody into the state space equation of the discrete time system;
2.2 Euclidean distance between antibodies was used to calculate affinity:
wherein abi,kDenotes the k dimension, ab of antibody ij,kDenotes the kth dimension of antibody j, L denotes the total antibody-encoding dimension, L ═ 6;
2.3 antibody concentration calculation algorithm:
wherein N is the number of immunized individuals and is 80; s (ab)i,abj) Is the similarity of the antibodies:
wherein δ is 0.2;
2.4 calculate encouragement values from affinity and antibody concentration:
sorting the encouragement values from high to low, and selecting the first 50% of the sorted individuals for reservation;
2.5 continuously generating antibodies of new generation for the reserved individuals through the actions of cloning, mutation and clone inhibition operators, reserving the first 50% of the individuals with the encouragement value from high to low in the antibodies generated in each generation, repeating the actions of cloning, mutation and clone inhibition operators to generate antibodies of the next generation until the immunization generation G reaches 400, reserving the individuals with the highest encouragement value obtained in the last generation to obtain each parameter matrix;
3) establishing a discrete sliding mode control law
The control method selects a sliding mode variable structure control method based on a discrete index approach law, and the selected switching surface is as follows:
s(k)＝E-Cx(k) (3.6)
wherein the coefficient matrix of the switching plane is C ═ 5, 1; e represents a target state value, and defines a switching zone including a switching surface on the basis of the switching surface s (k):
SΔ＝{x∈Rn|-Δ＜s(k)＜+Δ} (3.7)
2 Δ is the width of the switching band, Δ is 0.01; when determining the control law u (k) in the state space equation of the discrete variable structure system, firstly defining the form of discrete exponential approximation law:
s(k+1)-s(k)＝-qTs(k)-εTsat(s(k)),ε＞0,q＞0 (3.8)
wherein T is the sampling period;
according to the equation of state space, the switching surface and the discrete index approach law, the discrete sliding mode control law is obtained as follows:
u(k)＝-(Cb)-1[CAx(k)-(1-qT)s(k)+εTsat(s)] (3.10)
in the above formula, sat(s) is a saturation function, which is used to replace the sign function in the original approach law, and the function is:
step four: the method comprises the following steps of (1) searching the optimal values of the coefficients q and epsilon of the discrete sliding mode control law by using a genetic algorithm to obtain the discrete sliding mode control law, wherein the specific process comprises the following steps:
4.1 substituting the parameters A and b of the state space equation obtained in the step three into a discrete sliding mode control law; setting the value ranges of the discrete sliding mode control law coefficients q and epsilon;
4.2, coding two unknown discrete sliding mode control law coefficients q and epsilon, and converting decimal numbers into binary character strings, wherein the length of the binary character strings is 9;
4.3 importing the state value x (1) of the experimental data of the transient process at the first moment acquired in the step two, substituting x (1) and a group of initial q and epsilon into a discrete sliding mode control law to obtain u (1), substituting u (1) into a state space equation, and calculating the system state x at the second moment of simulationpro(2) By calculating sequentially, the simulated state value x at all times is obtainedpro(3),xpro(4),...,xpro(K) Wherein x ispro(k) Is the analog state value at the moment of K, K is more than or equal to K>1; x to be simulatedpro(k) The difference with the target state value E is recorded as r, and the target functions objvsc of the discrete sliding mode control law coefficients q and epsilon are as follows:
objvsc＝r＝∑[E-xpro(k)] (4.1)
the control purpose of the dust remover system is to reduce the value of the objective function objvsc, and to set the fitness function Fit to be the same as the objective function objvsc; the value of the objective function objvsc calculated by each group q and epsilon is the value of the fitness function Fit corresponding to a group q and epsilon;
4.4 setting the initial individual number to be 40, and adopting a random traversal sampling algorithm by a selection operator in the genetic algorithm:
where Fit (x)i) For each individual fitness value, F (x)i) Is an individual xiThe selected probability value N is the total number of the individuals;
4.5 selecting a single-point crossing algorithm, wherein the crossing probability is 0.6; selecting an algorithm of basic bit variation, wherein the variation probability is 0.05; carrying out cross variation on the initial 40 individuals to obtain a new generation of individuals, setting the number of genetic iteration individuals to be 50, selecting 50 individuals with higher selected probability to carry out cross variation to generate a next generation of individuals once in each genetic iteration, and continuously iterating until the number of the genetic iteration reaches 100; outputting the individuals with the highest selected probability in the individuals generated by the last genetic iteration, decoding the individuals according to the step 4.2, and calculating to obtain values of q and epsilon;
4.6 substituting the values of q and epsilon obtained in the step 4.5 into a discrete sliding mode control law to obtain the discrete sliding mode control law without unknown coefficients;
step five: starting the bag type dust collector system, and adjusting the frequency of a frequency converter of the bag type dust collector system and the opening of an electric control valve to be the value of an input variable corresponding to the initial state value of the system; according to a set target state value E of the bag type dust collector system, a switching surface and a discrete sliding mode control law without unknown coefficients, the frequency and the opening degree of an electric control valve of a frequency converter to be achieved by first regulation are calculated by combining with a system initial state value, the system state value is monitored, the system state value is collected once at intervals of a certain time, the frequency and the opening degree of the electric control valve of the frequency converter to be achieved by next regulation are calculated according to the system state value, the frequency and the opening degree of the electric control valve of the frequency converter are continuously monitored and calculated, the frequency and the opening degree of the electric control valve of the frequency converter are continuously regulated and controlled until the system state value reaches the target state value E, and accordingly, the control of the bag type dust collector system is completed.
2. The method according to claim 1, wherein the step one of obtaining the steady-state mathematical model of the output variables comprises the following steps:
1) selecting 16 induced draft fan frequency converters with the value range of 15 Hz-60 Hz, and selecting 12 electric control valves with the opening range of 5% -60%; expressing the frequency of the frequency converter as XiThe opening of the electrically controlled valve is represented as Yj(ii) a The specific combination form of the selected frequency of the frequency converter and the opening of the electric control valve is as follows:
sequentially combining the frequency of each frequency converter and the numerical value of the opening of the electric control valve to obtain 192 combined input conditions in total;
2) initializing a parameter variable of a dust remover system; starting a dust remover system, adjusting the frequency of a frequency converter to 15Hz, adjusting the opening of an electric control valve to 5%, monitoring the value of an output variable of the dust remover system at the moment, and when the value of the output variable is stable and does not change, taking the value of the output variable at the moment as the initial state value of the system; the airflow monitored to obtain the initial state value of the system is 425m3·h-1The air pressure difference of the air purifying chamber is-342 Pa;
3) sequentially adjusting input variables of a dust remover system according to the 192 combination input conditions obtained in the step 1), recording the value of the output variable corresponding to each combination input condition, analyzing and fitting the data of the two input variables and the two output variables, and constructing a steady-state mathematical model;
the steady state mathematical model of the airflow rate is:
the steady state mathematical model of the net air chamber pressure difference is as follows:
in the above two steady-state mathematical models, xbRepresenting the frequency of the frequency converter, xdIndicating the opening of the electrically controlled valve, F (x)b,xd) Represents the flow rate of the air flow in m3·h-1，P(xb,xd) Representing the net air chamber pressure difference in Pa.
3. The method of claim 1, wherein the target state value of the system in step two is set to 535m3·h-1The air pressure difference of the air purifying chamber is-478 Pa, the frequency of the corresponding target frequency converter is 30Hz, and the opening degree of the target electric control valve is 40%.
4. The method according to claim 1, wherein the ranges of the elements of the parameter A, b in step three are all set to [ -2,2], and the corresponding obtained parameter matrices are:
5. the method for designing the control strategy of the bag type dust collector according to claim 1, wherein in the fourth step, the value range of q is set as [8,12], the value range of epsilon is set as [0,1], and q is 10 and epsilon is 0.35 through corresponding calculation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110038227.5A CN112882387B (en) | 2021-01-12 | 2021-01-12 | Control strategy design method of bag type dust collector |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110038227.5A CN112882387B (en) | 2021-01-12 | 2021-01-12 | Control strategy design method of bag type dust collector |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112882387A true CN112882387A (en) | 2021-06-01 |
| CN112882387B CN112882387B (en) | 2022-06-03 |
Family
ID=76044647
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110038227.5A Active CN112882387B (en) | 2021-01-12 | 2021-01-12 | Control strategy design method of bag type dust collector |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112882387B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113521899A (en) * | 2021-07-29 | 2021-10-22 | 福建龙净环保股份有限公司 | Ash removal method of bag type dust collector |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE10013541A1 (en) * | 2000-03-20 | 2001-10-04 | Univ Halle Wittenberg | Process adaptive optimization system for industrial communication system, changes system parameters based on communication system characters measured according to data traffic |
| CN105116732A (en) * | 2015-09-01 | 2015-12-02 | 中冶华天工程技术有限公司 | Time-window-enthalpy-value-integral-based smoke cooling control method and system |
| CN109660162A (en) * | 2018-04-28 | 2019-04-19 | 西安中车永电电气有限公司 | A kind of immune sliding-mode control suitable for induction machine |
| CN110907835A (en) * | 2019-12-10 | 2020-03-24 | 北京理工大学 | Battery model parameter identification and SOC estimation method with noise immunity characteristic |
-
2021
- 2021-01-12 CN CN202110038227.5A patent/CN112882387B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE10013541A1 (en) * | 2000-03-20 | 2001-10-04 | Univ Halle Wittenberg | Process adaptive optimization system for industrial communication system, changes system parameters based on communication system characters measured according to data traffic |
| CN105116732A (en) * | 2015-09-01 | 2015-12-02 | 中冶华天工程技术有限公司 | Time-window-enthalpy-value-integral-based smoke cooling control method and system |
| CN109660162A (en) * | 2018-04-28 | 2019-04-19 | 西安中车永电电气有限公司 | A kind of immune sliding-mode control suitable for induction machine |
| CN110907835A (en) * | 2019-12-10 | 2020-03-24 | 北京理工大学 | Battery model parameter identification and SOC estimation method with noise immunity characteristic |
Non-Patent Citations (5)
| Title |
|---|
| LI Y,ET AL.: "Adaptive fuzzy neural sliding mode control based on immune genetic algorithm for ammunition auto-loading manipulator", 《JOURNAL OF COMPUTATIONAL INFORMATION SYSTEMS》, 31 January 2013 (2013-01-31), pages 6801 - 6810 * |
| 吴建辉等: "基于改进粒子群优化算法的离散混沌系统神经滑模控制", 《控制与决策》, no. 07, 15 July 2013 (2013-07-15), pages 137 - 141 * |
| 寇攀高: "水轮发电机及其调速系统的参数辨识方法与控制策略研究", 《中国博士学位论文全文数据库 工程科技II辑》, no. 9, 15 September 2012 (2012-09-15), pages 037 - 12 * |
| 朱晓瑾 等: "超超临界1000 MW燃煤机组干式电除尘器特性分析", 《热力发电》, vol. 49, no. 2, 29 November 2019 (2019-11-29), pages 99 - 103 * |
| 李晓斌: "真空退火炉智能变结构控制方法的研究与应用", 《材料热处理学报》, 25 February 2006 (2006-02-25), pages 124 - 129 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113521899A (en) * | 2021-07-29 | 2021-10-22 | 福建龙净环保股份有限公司 | Ash removal method of bag type dust collector |
| CN113521899B (en) * | 2021-07-29 | 2023-01-03 | 福建龙净环保股份有限公司 | Ash removal method of bag type dust collector |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112882387B (en) | 2022-06-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102175345B (en) | Soft measurement method for fire box temperature of multi-nozzle opposed coal water slurry gasification furnace | |
| KR102018330B1 (en) | Apparatus and method for detecting anomaly behavior in plant pipe using multiple meta-learning | |
| CN103853106A (en) | Energy consumption prediction parameter optimization method of building energy supply device | |
| CN112882387B (en) | Control strategy design method of bag type dust collector | |
| CN108760592B (en) | Fly ash carbon content online measurement method based on BP neural network | |
| CN109615082A (en) | It is a kind of based on stack selective ensemble learner air in fine particle PM2.5The prediction technique of concentration | |
| CN114225662B (en) | Hysteresis model-based flue gas desulfurization and denitrification optimal control method | |
| CN112446091A (en) | Artificial neural network-based pulsating pressure prediction method | |
| KR102204724B1 (en) | Apparatus and method for controlling OWC(Oscillating Water Column)-type wave power generator system using digital twin technology | |
| CN102135021A (en) | Method for predicting shaft power of industrial extraction condensing steam turbine | |
| CN107782857A (en) | Flexible measurement method based on the edible fungus fermented process key parameter for improving CS BPNN | |
| CN109492319A (en) | A kind of power plant boiler flue gas oxygen content flexible measurement method | |
| CN115510904A (en) | Boiler heating surface ash deposition monitoring method based on time sequence prediction | |
| CN111275204B (en) | Transformer state identification method based on hybrid sampling and ensemble learning | |
| CN112926767A (en) | Annular fog flow gas phase apparent flow velocity prediction method based on particle swarm BP neural network | |
| CN114113773A (en) | Non-invasive load monitoring method based on zero sample learning | |
| CN108537357A (en) | Photovoltaic power generation quantity loss forecasting method based on derating factor | |
| CN110737198A (en) | Large-scale coal-fired power plant CO based on BP neural network2Capture system prediction control method | |
| CN108897221B (en) | Military power supply chassis fan rotating speed control method | |
| CN108803750A (en) | A kind of forming machine high temperature humidity self-control energy saver and its processing method | |
| CN109376903A (en) | A kind of PM2.5 concentration value prediction technique based on game neural network | |
| CN112365082A (en) | Public energy consumption prediction method based on machine learning | |
| CN116956156A (en) | Sewage treatment intelligent aeration method and system based on artificial intelligence | |
| CN113095534A (en) | Wind power prediction method combining ARIMA and improved Elman neural network | |
| CN112801388B (en) | Power load prediction method and system based on nonlinear time series algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |