CN115453863A - Gas turbine load control method and device based on feedforward control of support vector machine - Google Patents
Gas turbine load control method and device based on feedforward control of support vector machine Download PDFInfo
- Publication number
- CN115453863A CN115453863A CN202210985677.XA CN202210985677A CN115453863A CN 115453863 A CN115453863 A CN 115453863A CN 202210985677 A CN202210985677 A CN 202210985677A CN 115453863 A CN115453863 A CN 115453863A
- Authority
- CN
- China
- Prior art keywords
- support vector
- vector machine
- gas turbine
- valve
- load
- 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.)
- Pending
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
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E20/00—Combustion technologies with mitigation potential
- Y02E20/16—Combined cycle power plant [CCPP], or combined cycle gas turbine [CCGT]
Abstract
The invention discloses a gas turbine load control method and a gas turbine load control device based on feedforward control of a support vector machine, wherein the method comprises the following steps: acquiring parameters of a gas turbine system, and processing the parameters by using a pre-trained support vector machine model to obtain a command of a pre-mixing valve; carrying out PID (proportion integration differentiation) operation on the deviation between the actual load and the set load value of the unit to obtain an adjustment quantity; superposing the instruction of the premixing valve as a feedforward value with the regulating quantity to obtain the opening of the premixing valve; and realizing the load control of the gas turbine based on the opening degree of the pre-mixing valve. According to the invention, after PID operation is carried out on the deviation of the actual load of the unit and the set value, the deviation and the feedforward value of the support vector machine are superposed to calculate the opening degree of the actuating mechanism, so that the lifting load control is carried out, and the defects of slow response, over-regulation, long stabilization time and the like of the traditional PID control system can be well overcome.
Description
Technical Field
The invention relates to the technical field of gas turbine control, in particular to a gas turbine load control method and device based on support vector machine feedforward control.
Background
The process flow of the gas-steam combined cycle unit is as follows: the natural gas enters a combustion chamber through a shut-off valve and a premixing valve, an air compressor continuously sucks air from the atmosphere and compresses the air, the compressed air enters the combustion chamber, is mixed with injected fuel and then is combusted into high-temperature gas, then the high-temperature gas flows into a gas turbine to complete expansion work, a turbine impeller is driven to rotate together, a generator is directly driven to generate power, the high-temperature flue gas after work is sent to a waste heat boiler, the heat of the high-temperature flue gas can generate high-temperature high-pressure steam, and then the generator set is driven to generate power, so that efficient gas-steam combined cycle is formed.
When the gas-steam combined cycle generator set normally operates, the steam turbine generally adopts a sliding pressure operation mode, the main steam regulating valve is fully opened, and the AGC function of the gas-steam combined cycle generator set is mainly undertaken by the gas turbine.
When the unit operates coordinately, the load instruction generation process of the gas turbine is as follows: and dispatching and issuing an AGC load instruction as a unit total load instruction, limiting the unit high and low load and limiting the unit lifting rate, sending the unit total load instruction to a gas turbine for setting a load instruction, and calculating the deviation of the load instruction and the actual load by a load controller to obtain the opening of a fuel valve (a premix valve) of the gas turbine. The method comprises the following steps that a unit total load instruction link comprises coordinated input/removal operation, when coordinated input is carried out, the unit total load instruction is sent to a combustion engine side through load and rate limitation to serve as a set value, and otherwise, the unit total load instruction tracks the actual total load of a unit.
Generally, the main adjustment target of the load regulator is the deviation between the load set value and the actual load. Meanwhile, according to the principle that power and frequency are consistent, load control can be achieved through rotating speed deviation, a primary frequency modulation loop of the unit responding to the power grid frequency deviation converts the frequency deviation into load deviation and superposes the load deviation on the input end of the load regulator, and the frequency modulation purpose is achieved. Because the static characteristic of the gas turbine determines that the power and the frequency have corresponding relation, the rotating speed and the load can be converted mutually through the rotating speed unequal rate, and can be actually regarded as the same regulated quantity. Thus, in a gas turbine control system, speed and load control share a single PID regulator, and the principle of the load control loop is shown in FIG. 1.
Since the gas turbine is a multivariable nonlinear delay system, the conventional PID control has certain defects, such as excessive overshoot, long time for adjusting the fuel quantity to be stable, and troublesome PID parameter adjustment.
In the related art, chinese patent publication No. CN111159844a describes a method for detecting an abnormality in the exhaust temperature of a gas turbine in a power station, which includes the following steps: selecting relevant parameters of a model; (2) collecting data samples; (3) data steady-state screening; (4) data normalization processing; (5) Selecting, training and testing a least square support vector machine model; (6) acquiring and statistically analyzing exhaust temperature residuals; and (7) detecting, analyzing and judging the exhaust temperature abnormity. But the scheme realizes accurate monitoring of the exhaust temperature abnormity of the gas turbine of the power station by means of a least square support vector machine model, and not realizes load control of the gas turbine.
Disclosure of Invention
The invention aims to solve the technical problem of how to quickly and stably adjust the fuel quantity and reduce the overshoot.
The invention solves the technical problems through the following technical means:
the invention provides a gas turbine load control method based on feedforward control of a support vector machine, which comprises the following steps:
acquiring parameters of a gas turbine system, wherein the parameters comprise a load instruction, an IGV opening degree, a natural gas temperature, a natural gas pressure, a natural gas heat value and an environment temperature;
processing the parameters by utilizing a feedforward control loop containing a pre-trained support vector machine model to obtain a command of the premixing valve;
carrying out PID (proportion integration differentiation) operation on the deviation between the actual load and the set load value of the unit to obtain an adjustment quantity;
superposing the instruction of the premixing valve as a feedforward value with the regulating quantity to obtain the opening of the premixing valve;
and realizing the load control of the gas turbine based on the opening degree of the pre-mixing valve.
The invention utilizes the technology of combining mechanism analysis, advanced algorithm and control theory, considers the change of load, natural gas parameters and environment variables, optimizes the load control system of the gas turbine, and after the deviation of the actual load of the unit and the set value is calculated by PID, the deviation and the feedforward value of the support vector machine are superposed to calculate the opening degree of the actuating mechanism, thereby carrying out the lifting load control; compared with the existing system, the gas turbine load control system based on the feedforward control of the support vector machine can well overcome the defects of slow response, over-regulation, long stabilization time and the like of the traditional PID control system; and the control structure based on the feedforward of the support vector machine has high accuracy of the load control of the gas turbine, is very suitable for being applied to engineering practice, fully improves the reliability and the availability of the gas turbine, prolongs the service life of the gas turbine to the maximum extent, and reduces the operation and maintenance cost.
Further, the method further comprises:
from gas turbine generator set history stationsSet historical data to construct data samples (x) i ,y i ) Wherein x is i As an input variable, y i To an output target value, the input variable comprising a parameter of the gas turbine system, the output target value being the premix valve command;
and training a least square support vector machine model based on the sample data to obtain the trained support vector machine model.
Further, the least squares support vector machine model is:
in the formula: k (x) i X) is a kernel function, a polynomial kernel function is used, where K (x) i ,x)＝(ax t y+c) d A is a support vector coefficient, and c is a constant offset; alpha is alpha i ，α i * Lagrange factors, respectively; b is a threshold value.
Further, the feedforward control loop including the pre-trained support vector machine model includes: the method comprises the following steps of configuring an input register, a support vector machine operation module, an element mapping register and an output register, wherein a pre-trained support vector machine model is deployed in the support vector machine operation module, and processing parameters by using a feedforward loop containing the pre-trained support vector machine model to obtain a premix valve instruction, and the method comprises the following steps:
writing a current signal containing parameters of the gas turbine system to the input register and constantly refreshed by the input register;
reading the current signal by using the support vector machine operation module, and obtaining an operation result by using the support vector machine model;
and writing the operation result into the element mapping register, transferring the stored state of the operation result to the output latch by the element mapping register, and outputting a current signal of a premix valve instruction by the output latch.
Further, the calculating the premix valve command as a feed-forward value and the adjustment amount in a superposition manner to obtain the opening degree of the premix valve specifically includes:
and superposing the premixed valve instruction as a feedforward value A and the PID calculated adjustment quantity B to obtain the opening C of the premixed valve, wherein C = A + B.
In addition, the invention also provides a gas turbine load control device based on feedforward control of a support vector machine, which comprises: the load control loop comprises a PID controller and a second adder, the actual load and the set load of the unit are used as the input of the second adder, the output of the second adder is connected with the PID controller, the output of the PID controller and the output of the feedforward control loop are both connected with the first adder, and the output of the first adder is connected with a premixing valve;
the feedforward control loop is provided with a pre-trained support vector machine model, the input of the support vector machine model is a parameter of a gas turbine system, and the output of the support vector machine model is a premixed valve instruction, wherein the parameter comprises a load instruction, an IGV opening degree, a natural gas temperature, a natural gas pressure, a natural gas heat value and an environment temperature.
Further, the apparatus further comprises:
a sample construction module for collecting historical data construction data samples (x) from a history station of the gas turbine generator set i ,y i ) Wherein x is i As an input variable, y i To an output target value, the input variable comprising a parameter of the gas turbine system, the output target value being the premix valve command;
and the model training module is used for training the least square support vector machine model based on the sample data to obtain the trained support vector machine model.
Further, the least squares support vector machine model is:
in the formula: k (x) i X) is a kernel function, a polynomial kernel function is used, where K (x) i ,x)＝(ax t y+c) d A is a support vector coefficient, and c is a constant offset; alpha is alpha i ，α i * Lagrange factors, respectively; b is a threshold value.
Further, the feed forward control loop comprises: the device comprises an input register, a support vector machine operation module, an element mapping register and an output register, wherein a pre-trained support vector machine model is deployed in the support vector machine operation module, and the device comprises:
said input register for writing a current signal containing parameters of said gas turbine system and continuously refreshing said current signal;
the support vector machine operation module is used for reading the current signal from the input register, obtaining an operation result by using the support vector machine model and writing the operation result into the element mapping register;
and the element mapping register is used for transferring the stored state to the output latch, and the output latch outputs a current signal of a premix valve instruction.
Further, the first adder is configured to perform superposition calculation on the premixed valve command serving as a feed-forward value a and the post-PID-operation adjustment amount B to obtain an opening C of the premixed valve, where C = a + B.
Further, the premixing valve adopts a hydraulic actuator.
The invention has the advantages that:
(1) The invention utilizes the technology of combining mechanism analysis, advanced algorithm and control theory, considers the change of load, natural gas parameters and environment variables, optimizes the load control system of the gas turbine, and after the deviation of the actual load of the unit and the set value is calculated by PID, the deviation and the feedforward value of the support vector machine are superposed to calculate the opening degree of the actuating mechanism, thereby carrying out the lifting load control; compared with the existing system, the gas turbine load control system based on the feedforward control of the support vector machine can well overcome the defects of slow response, over-regulation, long stabilization time and the like of the traditional PID control system; and the control structure based on the feedforward of the support vector machine has high accuracy of the load control of the gas turbine, is very suitable for being applied to engineering practice, fully improves the reliability and the availability of the gas turbine, prolongs the service life of the gas turbine to the maximum extent, and reduces the operation and maintenance cost.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
FIG. 1 is a schematic diagram of a conventional load control loop as mentioned in the background of the invention section;
FIG. 2 is a schematic flow chart of a gas turbine load control method based on support vector machine feedforward control according to a first embodiment of the present invention;
FIG. 3 is a schematic diagram of the operating path of the feedforward control loop in a first embodiment of the invention;
FIG. 4 is a schematic diagram of a support vector machine model in a first embodiment of the invention;
FIG. 5 is a schematic diagram of a gas turbine load control apparatus based on support vector machine feed forward control according to a second embodiment of the present invention;
FIG. 6 is a block diagram of a model of the governor of the unit of the present invention;
FIG. 7 is a block diagram of the electro-hydraulic servo mechanism model of the present invention;
FIG. 8 is a block diagram of the construction of a prime mover model in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention, and it is obvious that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in FIG. 2, a first embodiment of the present invention provides a method for controlling the load of a gas turbine based on feedforward control of a support vector machine, the method comprising the following steps:
s10, obtaining parameters of a gas turbine system, wherein the parameters comprise a load instruction, an IGV opening degree, a natural gas temperature, a natural gas pressure, a natural gas heat value and an environment temperature;
s20, processing the parameters by utilizing a feedforward control loop containing a pre-trained support vector machine model to obtain a command of the pre-mixing valve;
s30, obtaining an adjustment quantity by performing PID (proportion integration differentiation) operation on the deviation between the actual load and the set load value of the unit;
s40, overlapping and calculating the instruction of the pre-mixing valve as a feed-forward value and the regulating quantity to obtain the opening degree of the pre-mixing valve;
and S50, realizing load control of the gas turbine based on the opening degree of the premixing valve.
The embodiment optimizes a load control system of the gas turbine, after the deviation of the actual load of the unit and a set value is subjected to PID operation, the deviation and a feedforward value of a support vector machine are subjected to superposition calculation to obtain the opening degree of an actuating mechanism, and therefore lifting load control is performed; compared with the existing system, the gas turbine load control system based on support vector machine feedforward control can well overcome the defects of slow response, over-regulation, long stabilization time and the like of the traditional PID control system by the support vector machine feedforward; and the control structure based on the feedforward of the support vector machine has high accuracy of the load control of the gas turbine, is very suitable for being applied to engineering practice, fully improves the reliability and the availability of the gas turbine, prolongs the service life of the gas turbine to the maximum extent, and reduces the operation and maintenance cost.
In an embodiment, the method further comprises the steps of:
construction of data sample (x) from historical data collected from gas turbine generator set historical station i ,y i ) Wherein x is i As input variables, x i ∈R n ，y i To output the target value, y i E.g. R, n is the total number of sample points,the input variables include parameters of the gas turbine system and the output target values are the premix valve commands;
and training a least square support vector machine model based on the sample data to obtain the trained support vector machine model.
It should be noted that training of SVM (support vector machine) is equivalent to solving quadratic programming problem with linear constraint to obtain global optimal solution. In addition, the topological structure is only related to the support vector, so that the calculation amount is reduced, and the online application is facilitated.
In this embodiment, to implement the gas turbine load control system based on the support vector machine feedforward control, the accuracy of the feedforward prediction value of the support vector machine needs to be ensured, so that the whole system is stable and does not oscillate. By introducing an insensitive loss function, the support vector machine is expanded to solve the problem of nonlinear regression estimation, and compared with a neural network method, the method avoids the problem of falling into local minimum and has obvious superiority.
In one embodiment, a non-linear mapping is assumed
Can be mapped from the input space to the output space and then passed through this mapping
The input space x is mapped to a high-dimensional feature space F and a linear regression is performed in this high-dimensional feature space using the function (1):
in the formula: b is a threshold value, and the coefficients w and b can be estimated by minimizing equation (2):
in the formula: l epsilon is a loss function;
the function is more flat due to the regularization part, so that the generalization capability of the estimation function is improved; c is a penalty parameter used to decide the balance between regularization component and empirical risk;
is an empirical risk among which are:
in the formula: ε is the insensitive loss function.
The support vector machine determines the regression function by minimizing the following objective function:
the constraint conditions are as follows:
introducing relaxation variable xi in formula i And xi i * The objective is to have a solution to the regression function.
In general, a Lagrange optimization method is introduced by establishing a Lagrange equation to solve a regression function equation, and meanwhile, the Lagrange equation is ordered
Solving to obtain the least square support vector machine model as follows:
in the formula: k (x) i X) is a kernel function, using a polynomial kernel functionNumber, where K (x) i ,x)＝(ax t y+c) d A is a support vector coefficient, and c is a constant offset; alpha is alpha i ，α i * Lagrange factors, respectively; b is a threshold value.
The polynomial kernel function in the embodiment is very suitable for data after orthogonal normalization, and has high stability.
In one embodiment, as shown in fig. 3 to 4, the feedforward control loop including the pre-trained support vector machine model includes: the method comprises the following steps of configuring an input register, a support vector machine operation module, an element mapping register and an output register, wherein a pre-trained support vector machine model is deployed in the support vector machine operation module, and processing parameters by using a feedforward loop containing the pre-trained support vector machine model to obtain a premix valve instruction, and the method comprises the following steps:
writing a current signal containing parameters of the gas turbine system to the input register and constantly refreshed by the input register;
reading the current signal by using the support vector machine operation module, and obtaining an operation result by using the support vector machine model;
and writing the operation result into the element mapping register, transferring the stored state of the operation result to the output latch by the element mapping register, and outputting a current signal of a premix valve instruction by the output latch.
It should be noted that, in this embodiment, the current signal of 4 to 20mA of the load instruction, the IGV opening, the natural gas temperature, the natural gas pressure, the natural gas heat value, and the ambient temperature on the site is collected and written into the input register, and at this time, the input register is continuously refreshed. In the support vector machine operation module, the operation module continuously executes the calculation program, and the calculation result is sent to the element mapping register. And transferring the state in the element mapping register to an output latch, and converting the state into a 4-20 mA current signal of a command of the premix valve through the isolation of an output module. And then superposing a 4-20 mA current signal of the instruction of the pre-mixing valve with the deviation between the actual load and the set load value of the unit through a PID operation result, and finely adjusting the instruction of the pre-mixing valve by using the PID operation result to obtain the opening of the pre-mixing valve. The defects of slow response, over-regulation, long stabilization time and the like of the traditional PID control system can be well overcome by utilizing the feedforward of the support vector machine.
In an embodiment, the step S40: and superposing the instruction of the pre-mixing valve as a feedforward value with the regulating quantity to obtain the opening degree of the pre-mixing valve, wherein the method specifically comprises the following steps: and superposing the premixed valve instruction as a feedforward value A and the PID calculated adjustment quantity B to obtain the opening C of the premixed valve, wherein C = A + B.
It should be noted that the instruction of the premix valve carries the opening of the premix valve, and the value range of the opening of the premix valve is 0 to 100%, in this embodiment, the feed-forward value of the support vector machine is mainly used, and the feed-forward value of the support vector machine is finely adjusted by using the result of PID operation of the deviation between the actual load of the unit and the set value, so as to obtain the final opening of the premix valve, thereby realizing the control of the lifting load.
Further, as shown in fig. 5, a second embodiment of the present invention provides a gas turbine load control apparatus based on support vector machine feedforward control, the apparatus including: the load control loop comprises a PID controller and a second adder, the actual load and the set load of the unit are used as the input of the second adder, the output of the second adder is connected with the PID controller, the output of the PID controller and the output of the feedforward control loop are both connected with the first adder, and the output of the first adder is connected with a premixing valve;
the feedforward control loop is provided with a pre-trained support vector machine model, the input of the support vector machine model is a parameter of a gas turbine system, and the output of the support vector machine model is a premixed valve instruction, wherein the parameter comprises a load instruction, an IGV opening degree, a natural gas temperature, a natural gas pressure, a natural gas heat value and an environment temperature.
The embodiment optimizes the load control system of the gas turbine, adds a feedforward control loop, utilizes the learning capability and universal approximation of the support vector machine to simulate the inverse system of the gas turbine, namely inputs the load instruction, the IGV opening, the natural gas temperature, the natural gas pressure, the natural gas heat value and the environment temperature; the output is a command of a pre-mixing valve; after the deviation between the actual load of the unit and the set value is subjected to PID operation, the deviation and the feedforward value of the support vector machine are subjected to superposition calculation to obtain the opening degree of an actuating mechanism, so that the lifting load is controlled; compared with the existing system, the gas turbine load control system based on support vector machine feedforward control can well overcome the defects of slow response, over-regulation, long stabilization time and the like of the traditional PID control system by the support vector machine feedforward; and the control structure based on the feedforward of the support vector machine has high accuracy of the load control of the gas turbine, is very suitable for being applied to engineering practice, fully improves the reliability and the availability of the gas turbine, prolongs the service life of the gas turbine to the maximum extent, and reduces the operation and maintenance cost.
In one embodiment, the apparatus further comprises:
a sample construction module for collecting historical data from a gas turbine generator set historical station to construct a data sample (x) i ,y i ) Wherein x is i As an input variable, y i To an output target value, the input variables comprising parameters of the gas turbine system, the output target value being the premix valve command;
and the model training module is used for training the least square support vector machine model based on the sample data to obtain the trained support vector machine model.
In one embodiment, the least squares support vector machine model is:
in the formula: k (x) i X) is a kernel function, a polynomial kernel function is used, where K (x) i ,x)＝(ax t y+c) d A is a support vector coefficient, and c is a constant offset; alpha is alpha i ，α i * Lagrange factors, respectively; b is a threshold value.
The polynomial kernel function adopted by the embodiment is very suitable for data after orthogonal normalization, and has high stability.
In one embodiment, further, the feed forward control loop comprises: the device comprises an input register, a support vector machine operation module, an element mapping register and an output register, wherein a pre-trained support vector machine model is deployed in the support vector machine operation module, and the device comprises:
said input register for writing a current signal containing parameters of said gas turbine system and continuously refreshing said current signal;
the support vector machine operation module is used for reading the current signal from the input register, obtaining an operation result by using the support vector machine model and writing the operation result into the element mapping register;
and the element mapping register is used for transferring the stored state to the output latch, and the output latch outputs a current signal of a premix valve instruction.
In an embodiment, the first adder is configured to calculate the premix valve command as a feed-forward value a and the post-PID-operation adjustment amount B in a superimposed manner, so as to obtain an opening C of the premix valve, where C = a + B.
In one embodiment, the premix valve employs a hydraulically actuated actuator.
Particularly, the hydraulic actuator moves up and down by the aid of the circular valve core to control the valve to be opened and closed, has good deviation resistance, can provide high thrust, runs very stably, has high response speed, and can realize high-precision control.
It should be noted that, other embodiments or implementation methods of the gas turbine load control device based on support vector machine feedforward control according to the present invention can refer to the above-mentioned embodiments, and no redundancy is necessary here.
The control effect of the gas turbine load control scheme based on the feedforward control of the support vector machine proposed in the embodiment of the present specification is verified as follows:
(1) Model of a control system
The model of the speed regulator of the unit conforms to a GJ/GJ + card in a stability analysis program BPA, and a structural block diagram of the speed regulator of the unit is shown in a following figure 6. According to the measurement result, the pure DELAY time DELAY1=0.0133s of the frequency input signal of the fuel power measurement link is calculated, the time constant TR =0.11s of the first-order inertia link corresponding to the power feedback signal, and the PCV is a gate regulation instruction.
Through engineering empirical calibration, the load PID controller parameters are finally set to KP1=0.03, kd1=0, ki1=0.143.
(2) Electrohydraulic servo mechanism model
A GA card model in BPA can be adopted in an electro-hydraulic actuator model of a No. 2 unit of a certain power plant, as shown in figure 7, when a throttle instruction is greatly stepped, due to the amplitude limiting effect in a closed loop, the opening degree of the throttle is uniformly changed along with time at the initial stage of action, namely, the action speed of the servomotor is controlled by a nonlinear link, when the instruction and a feedback difference value are smaller, the throttle is in a state of jointly adjusting PID and servomotor characteristics, the nonlinear links such as amplitude limiting and the like do not work, and therefore the throttle characteristic can be equivalent to the following standard mathematical model:
wherein:
or
T c Is the closing time constant, T, of the servomotor o Is the starting time constant of the servomotor, P GV (s) premix valve opening feedback, P CV (s) is a premix valve opening command.
Therefore, the time constant of the gate adjusting characteristic can be obtained through the large-amplitude step and small-amplitude step instructions of the gate adjusting instruction and feedback calculation. Because the PID amplification factor of the throttle is generally a constant value, the value is equal when the throttle is opened and closed, in order to ensure the reasonability of the selected parameters, kp =10 is taken, and according to 50% -55% and 55% -50% step data in a field test, a standard throttle model is adopted, and characteristic parameters of each throttle are obtained through identification calculation and are shown in table 1, wherein the time constant of the starting of the servomotor is 0.38s, the time constant of the closing of the servomotor is 0.36s, and PGV is the opening degree of the throttle.
TABLE 1 actuator parameter identification results
(3) Prime mover model
The gas turbine has only one cylinder, the total load of the unit is mainly adjusted by the gas turbine, and the unit belongs to a reheater-free type, wherein a reheater-free turbine model (TA) is adopted as a prime mover model, and is shown in fig. 8.
The prime motor model adopts a TA card, mainly considers the dynamic characteristic of active power change caused by the opening of the premixing valve, and is a first-order inertia link. According to the fuel valve opening feedback and the unit power response curve at the disturbance moment of the disturbance test, the self response time constant of the combustion engine can be identified by applying identification software.
And identifying by a least square method according to the dynamic disturbance test data to obtain the volume time constant of the gas turbine, wherein the TCH average value is 5.0s, and PM is real-time active power.
Through verification, compared with the existing system, the gas turbine load control system based on the feedforward control of the support vector machine provided by the embodiment well overcomes the defects of slow response, overshoot, long stabilization time and the like of the traditional PID control system.
It should be noted that the logic and/or steps represented in the flowcharts or otherwise described herein, such as an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Further, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, e.g., two, three, etc., unless explicitly specified otherwise.
Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (10)
1. A method for controlling the load of a gas turbine based on support vector machine feed forward control, the method comprising:
acquiring parameters of a gas turbine system, wherein the parameters comprise a load instruction, an IGV opening degree, a natural gas temperature, a natural gas pressure, a natural gas heat value and an environment temperature;
processing the parameters by utilizing a feedforward control loop containing a pre-trained support vector machine model to obtain a command of the premixing valve;
performing PID operation on the deviation between the actual load and the set load value of the unit to obtain an adjustment quantity;
superposing the instruction of the premixing valve as a feedforward value with the regulating quantity to obtain the opening of the premixing valve;
and realizing load control of the gas turbine based on the opening degree of the premixing valve.
2. A method for support vector machine feed forward control based gas turbine load control as claimed in claim 1, wherein said method further comprises:
collecting historical data from a history station of a gas turbine generator set to construct a data sample (x) i ,y i ) Wherein x is i As an input variable, y i To an output target value, the input variables comprising parameters of the gas turbine system, the output target value being the premix valve command;
and training a least square support vector machine model based on the sample data to obtain the trained support vector machine model.
3. A support vector machine feed forward control based gas turbine load control method as claimed in claim 2, wherein said least squares support vector machine model is:
in the formula: k (x) i X) is a kernel function, a polynomial kernel function is used, where K (x) i ,x)＝(ax t y+c) d A is a support vector coefficient, and c is a constant offset; alpha is alpha i ，α i * Lagrange factors, respectively; b is a threshold value.
4. A support vector machine feedforward control-based gas turbine load control method as claimed in claim 1, wherein the feedforward control loop including the pre-trained support vector machine model includes: the method comprises the following steps of configuring an input register, a support vector machine operation module, an element mapping register and an output register, wherein a pre-trained support vector machine model is deployed in the support vector machine operation module, and processing parameters by using a feedforward loop containing the pre-trained support vector machine model to obtain a premix valve instruction, and the method comprises the following steps:
writing a current signal containing parameters of the gas turbine system to the input register and constantly refreshed by the input register;
reading the current signal by using the support vector machine operation module, and obtaining an operation result by using the support vector machine model;
and writing the operation result into the element mapping register, transferring the stored state of the operation result to the output latch by the element mapping register, and outputting a current signal of a premix valve instruction by the output latch.
5. A method of support vector machine feedforward control based gas turbine load control as claimed in claim 1, wherein said calculating the premix valve command as a feedforward value to be superimposed with the manipulated variable to obtain the opening of the premix valve comprises:
and superposing the pre-mixing valve instruction serving as a feedforward value A and the PID operated regulating quantity B to obtain the opening C of the pre-mixing valve, wherein C = A + B.
6. A gas turbine load control apparatus based on support vector machine feed forward control, the apparatus comprising: the load control loop comprises a PID controller and a second adder, the actual load and the set load of the unit are used as the input of the second adder, the output of the second adder is connected with the PID controller, the output of the PID controller and the output of the feedforward control loop are both connected with the first adder, and the output of the first adder is connected with a premixing valve;
the feedforward control loop is provided with a pre-trained support vector machine model, the input of the support vector machine model is a parameter of the gas turbine system, the output of the support vector machine model is a premixing valve instruction, and the parameter comprises a load instruction, an IGV opening degree, a natural gas temperature, a natural gas pressure, a natural gas heat value and an environment temperature.
7. A support vector machine feed forward control based gas turbine load control apparatus as claimed in claim 6, wherein said apparatus further comprises:
a sample construction module for collecting from a history station of the gas turbine generator setSet historical data to construct data samples (x) i ,y i ) Wherein x is i As an input variable, y i To an output target value, the input variables comprising parameters of the gas turbine system, the output target value being the premix valve command;
and the model training module is used for training the least square support vector machine model based on the sample data to obtain the trained support vector machine model.
8. A support vector machine feed forward control based gas turbine load control apparatus as claimed in claim 7 wherein said least squares support vector machine model is:
in the formula: k (x) i X) is a kernel function, a polynomial kernel function is used, where K (x) i ,x)＝(ax t y+c) d A is a support vector coefficient, and c is a constant offset; alpha is alpha i ，α i * Lagrange factors, respectively; b is a threshold value.
9. A support vector machine based feed forward control gas turbine load control apparatus as claimed in claim 6 wherein said feed forward control loop comprises: the device comprises an input register, a support vector machine operation module, an element mapping register and an output register, wherein a pre-trained support vector machine model is deployed in the support vector machine operation module, and the device comprises:
said input register for writing a current signal containing parameters of said gas turbine system and continuously refreshing said current signal;
the support vector machine operation module is used for reading the current signal from the input register, obtaining an operation result by using the support vector machine model and writing the operation result into the element mapping register;
and the element mapping register is used for transferring the stored state to the output latch, and the output latch outputs a current signal of a premix valve instruction.
10. A support vector machine feedforward control-based gas turbine load control device in accordance with claim 6, wherein the first adder is configured to add the premixed valve command as a feedforward value a to the PID-operated manipulated variable B to obtain an opening C of the premixed valve, where C = a + B.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210985677.XA CN115453863A (en) | 2022-08-17 | 2022-08-17 | Gas turbine load control method and device based on feedforward control of support vector machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210985677.XA CN115453863A (en) | 2022-08-17 | 2022-08-17 | Gas turbine load control method and device based on feedforward control of support vector machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115453863A true CN115453863A (en) | 2022-12-09 |
Family
ID=84297715
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210985677.XA Pending CN115453863A (en) | 2022-08-17 | 2022-08-17 | Gas turbine load control method and device based on feedforward control of support vector machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115453863A (en) |
-
2022
- 2022-08-17 CN CN202210985677.XA patent/CN115453863A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10732595B2 (en) | Control parameter automatic-adjustment apparatus, control parameter automatic-adjustment method, and control parameter automatic-adjustment apparatus network | |
| US5305230A (en) | Process control system and power plant process control system | |
| Wu et al. | Gain scheduling design based on active disturbance rejection control for thermal power plant under full operating conditions | |
| US11232376B2 (en) | System and method for optimizing combustion of boiler | |
| US20070055392A1 (en) | Method and system for model predictive control of a power plant | |
| CN106968803B (en) | Combined probability control of power output-emission parameters in gas turbine regulation | |
| CN106884725B (en) | Probabilistic control of power output-emission parameters in gas turbine regulation | |
| CN106884724B (en) | Probabilistic control modeling in gas turbine tuning of power output-emission parameters | |
| CN113031565A (en) | Method and system for predicting primary frequency modulation power response performance of thermal power generating unit | |
| CN106884726B (en) | System and method for regulation and control of a gas turbine | |
| CN106884727B (en) | Probabilistic control of power output-emission parameters in gas turbine regulation | |
| CN106988893B (en) | Probabilistic control of power output-emission parameters in gas turbine regulation | |
| US11287126B2 (en) | System and method for controlling operation of boiler | |
| CN115453863A (en) | Gas turbine load control method and device based on feedforward control of support vector machine | |
| Undrill et al. | Modeling of combined cycle plants in grid simulation studies | |
| Biyanto et al. | Design plant-wide control to waste heat recovery generation on cement industry based HYSYS | |
| Ahamed et al. | Reinforcement learning controllers for automatic generation control in power systems having reheat units with GRC and dead-band | |
| US9599025B2 (en) | Application of probabilistic control in gas turbine tuning for power output-exhaust energy parameters, related control systems, computer program products and methods | |
| US9599026B2 (en) | Application of probabilistic control in gas turbine tuning for exhaust energy-power output parameters, related control systems, computer program products and methods | |
| Vansovits et al. | Identification of industrial water boiler for model predictive control of district heat plant | |
| Jurado | Hammerstein‐model‐based predictive control of micro‐turbines | |
| Mukarromah et al. | Modeling of Level Control System using IMC-PID Tuning Method with Filter Tuning Factor Variation on Steam Drum Boiler of Power Plant | |
| Yulianto et al. | SYSTEM DEVELOPMENT FOR ENHANCING BOILER PERFORMANCE: FROM PID CONTROL TO ADAPTIVE AND MODEL PREDICTIVE CONTROL FOR MORE EFFECTIVE OPTIMIZATION OF TEMPERATURE, PRESSURE, AND LEVEL CONTROL IN BOILER SYSTEM | |
| Lindberg | A comparison between MPC and PID controllers for education and steam reformers | |
| Barghandan et al. | Design of a Model Reference Adaptive Controller Using Modified MIT Rule for a Second Order System |
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 |