WO2020062806A1 - 一种用于燃烧后co 2捕集系统的改进ina前馈控制方法 - Google Patents

一种用于燃烧后co 2捕集系统的改进ina前馈控制方法 Download PDF

Info

Publication number
WO2020062806A1
WO2020062806A1 PCT/CN2019/079091 CN2019079091W WO2020062806A1 WO 2020062806 A1 WO2020062806 A1 WO 2020062806A1 CN 2019079091 W CN2019079091 W CN 2019079091W WO 2020062806 A1 WO2020062806 A1 WO 2020062806A1
Authority
WO
WIPO (PCT)
Prior art keywords
particle
feedforward
ina
combustion
post
Prior art date
Application number
PCT/CN2019/079091
Other languages
English (en)
French (fr)
Inventor
沈炯
唐炜洁
张俊礼
吴啸
Original Assignee
东南大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 东南大学 filed Critical 东南大学
Publication of WO2020062806A1 publication Critical patent/WO2020062806A1/zh

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

Definitions

  • the invention belongs to the technical field of thermal control technology, and particularly relates to an improved INA feedforward control method for a post-combustion CO 2 capture system.
  • the post-combustion CO 2 capture system is based on the chemical absorption method and directly separates CO 2 from the flue gas after power plant combustion. It is the mainstream technology used in current CO 2 capture power plants in recent years. Flexible controllability is one of the requirements for the design and control of CO 2 capture systems. In the current research, the most extensive control method is the PID regulator control.
  • Control Strategy for CO 2 Capture System in Coal-fired Power Plant Flue Gas published by Niu Hongwei, etc., proposed the use of PID to control the recycle flow of the regeneration tower solution, coordinated control performed absorber liquid level, in order to achieve the purpose of safe and efficient;
  • LinY J et published “by using an absorption solution and monoethanolamine to widely control the pumping of CO 2 capture (Plant wide control of CO 2 capture by absorption in addition and stripping using monoethanolamine solution), a decentralized PI control scheme is proposed to maintain the CO 2 capture rate and reboiler temperature by controlling the reboiler temperature and lean liquid flow.
  • Chinese patent application 201710795146.3 discloses "a multi-model predictive control method for a post-combustion CO 2 capture system".
  • the predictive control method uses a chemical adsorption-based post-combustion CO 2 capture system as a controlled object, and the lean liquid valve is opened.
  • the opening degree of the extraction valve of the low-pressure cylinder of the steam turbine is the system control input, and the capture rate CO 2 and the reboiler temperature are the system output.
  • the data generated by the system operation is used at different operating conditions.
  • the multi-model method is used and multiple predictive controllers are applied, which makes the operation complicated and the amount of calculations large, which is not conducive to project implementation.
  • the PID method is usually used to solve, and the predictive control method using membership weighting is difficult to achieve in engineering, and it is easy to jump when the working conditions are changed.
  • Cida patent application 201711138022.4 discloses "a proxy model modeling method for coal water slurry gasification process". This method selects several measurable process states as input variables, including oxygen-coal ratio, coal slurry concentration, coal slurry flow, H / C element molar ratio in coal, O / C element molar ratio in coal, and ash content in coal.
  • the improved particle swarm algorithm is limited to identifying model parameters and cannot be applied to a post-combustion CO 2 capture system.
  • the improved particle swarm algorithm requires data pre-processing in advance and is applied to offline calculations. If the object changes in the case of disturbance in the industry, the object model cannot be obtained in real time.
  • the purpose of the present invention is to provide an improved INA feedforward control method for a post-combustion CO 2 capture system in order to overcome the shortcomings of the prior art.
  • the invention After being used in a post-combustion CO 2 capture system, the invention can quickly and smoothly track the set value of the changed CO 2 capture rate, effectively suppress the influence of flue gas disturbance, and have good anti-interference ability.
  • the design of the improved INA method is performed by identifying the controlled and controlled quantities of the post-combustion CO 2 capture system.
  • the integrated design of the improved INA and feedforward controller is completed, and then the control of the CO 2 capture rate y 1 and the reboiler temperature y 2 is implemented.
  • the specific steps include the following:
  • Step 1 Identification of controlled and controlled quantities of the post-combustion CO 2 capture system: adding two controlled quantities of lean liquid flow u 1 and turbine exhaust flow u 2 under steady-state conditions and flue gas disturbance
  • the quantity u 3 is the input step excitation signal data to obtain two controlled quantities of the CO 2 capture rate y 1 and the reboiler temperature y 2 as the output step response data.
  • the input and output data are identified and used as Controlled open-loop transfer function matrix G (s);
  • Step 2 Improve the design of the INA method: Use the improved INA method based on particle swarm optimization to decouple the controlled object, obtain the compensation matrix K p (s) through self-optimization, and form an equivalent object G (s) with the controlled object.
  • K p (s); the PID regulator K c (s) diag ⁇ k ci (s) ⁇ is set for the equivalent object, and the setting condition is to make the controlled amount described in step 1 have a smaller Overshoot, fast transition process, closed-loop steady-state error is zero, so as to meet the constraints and rate of change of the control amount described in step 1.
  • Step 3 Improve the integrated design of the INA and the feedforward controller: Based on the decoupled equivalent object, use the invariance principle method to design the feedforward controller to expand and improve the INA to suppress the influence of flue gas disturbances; set the feedforward The feedforward gain of the controller is used to adjust the size of the feedforward effect. If the physical form of the feedforward controller is not achievable, the inertia link is used instead.
  • the inertia link refers to Where T is the inertia time.
  • a further preferred solution of the improved INA feedforward control method for a post-combustion CO 2 capture system proposed by the present invention is:
  • the improved INA method based on particle swarm optimization in step 2 includes the following specific steps:
  • Step 21 Find the inverse of G (s), draw Gershgorin band, based on the principle that the Gershgorin band does not contain (0,0) points Whether it is a diagonal dominant array;
  • Step 23 Design a PID regulator k ci (s) for the equivalent object G (s) K p (s) to meet the requirements of transient and steady state response;
  • Step 24 Determine the feedback gain f according to the inverse Nai stability criterion, so that the closed-loop system is stable and has an appropriate stability margin;
  • step 25 the post-combustion CO 2 capture system is simulated. If it is not satisfied, return to step 22 to recompensate and readjust k ci (s) until it is satisfied.
  • the particle swarm algorithm described in step 22 includes the following specific steps:
  • Step 221 the form of the optimized compensation matrix is:
  • the particle swarm is composed of m particles, and each particle has 4 dimensions, which are a, b, c, and d variables, which are variables to be optimized;
  • Step 222 defining a performance function:
  • n is the number of frequency points
  • D point is the distance from the center of the circle to the origin (with the diagonal array g ii (jw) of G (s) as the center), and R is the corresponding radius;
  • the Gershgorin band can be moved away from the origin within the frequency point range, and the optimal value can be obtained within the frequency point range;
  • Step 223, update the particle position:
  • Equation (3) The inertia coefficient ⁇ is used to balance the global search capability and Local search capability, rand () is added to prevent falling into a local optimum,
  • Step 224 draw a Gershgorin band to verify whether G (s) K p (s) satisfies the diagonal advantage. If not, change the learning factor c 3 , the frequency point n, and the inertia weight ⁇ , and return to (222) for calculation.
  • step (3) The design of the feedforward controller using the principle of invariance described in step (3) means that the feedforward controller is added after the PID controller or for the constraints that need to deal with the output of the feedforward controller. Value; then the original coupled object is decomposed into two independent objects through improved INA method, then the transfer function equation in the feedforward controller is:
  • Y (s) represents the output, there are two components of CO 2 capture rate and reboiler temperature, X (s) represents the amount of flue gas input, i takes 1,2 to represent the decoupled
  • W o is the equivalent after compensation
  • W F is the feedforward controller
  • G di is the disturbance transfer function
  • f i is the loop gain
  • M (s), N (s) are the numerator and denominator of the achievable link in normal physics, the order of M (s) is lower than N (s), and a is the coefficient of the differential link;
  • the derivation link is physically unachievable, it is transformed into a first-order inertial link, which is equivalent at low frequencies:
  • T is a time constant.
  • the realization principle of the present invention is that the Inverse Nyquist Array (INA) referred to in the present invention is a control method of a linear multivariable system, although the INA control method is more effective in thermal energy control engineering than the existing PID method. It has advantages, but has not yet reached the optimal state of control. Therefore, an improved INA feedforward control method for a post-combustion CO 2 capture system proposed by the present invention is based on the post-combustion CO 2 capture system being controlled. Object, lean liquid flow and steam turbine extraction flow are controlled quantities, CO 2 capture rate and reboiler temperature are controlled quantities, and flue gas flow caused by power station power changes are changed to external disturbance quantities.
  • Object, lean liquid flow and steam turbine extraction flow are controlled quantities
  • CO 2 capture rate and reboiler temperature are controlled quantities
  • flue gas flow caused by power station power changes are changed to external disturbance quantities.
  • the improved INA method is used to decouple the CO 2 capture system, and then multivariable feedforward control is introduced for the decoupled equivalent object using the invariance principle method, so as to suppress the influence of the flue gas flow disturbance and avoid the need of the existing INA method. Tried and tested the shortcomings.
  • the improved INA feed-forward control method of the present invention can quickly and smoothly track the set value of the changed CO 2 capture rate after being used in a post-combustion CO 2 capture system, and effectively suppress the influence of flue gas disturbances, and has good anti-interference ability .
  • the present invention has significant advantages:
  • the present invention has good resistance to flue gas flow disturbance, and can quickly and flexibly track the set value of the CO 2 capture rate and keep the reboiler temperature near the set value.
  • the present invention can more effectively suppress the influence of flue gas disturbances and ensure the flexible operation of the CO 2 capture system after combustion; it is easy to implement in engineering, especially can be maintained under variable operating conditions Run smoothly.
  • the present invention uses the improved INA method based on particle swarm optimization to decouple the controlled object, so that it has self-optimizing ability, and solves the shortcomings of the existing INA that need to be tried out and the effect is not significant.
  • the present invention combines the improved INA method with a feedforward controller, which can achieve the suppression of measurable disturbances, so that the post-combustion CO 2 capture system can quickly and smoothly track the set value of the changed CO 2 capture rate, and is effective. Suppress flue gas disturbance and have good anti-interference ability.
  • the INA feedforward control method, the existing PID method and the existing INA method will be used to control the post-combustion CO 2 capture system respectively, and the control effect performance parameter table shown in Table 1 will be obtained.
  • the improved INA feedforward control method has a shorter adjustment transition process time, a smaller overshoot amount, and substantially no steady-state deviation, which illustrates the control method of the present invention. It can be applied to the post-combustion CO 2 capture system to obtain a good control effect.
  • Figure 1 is a schematic block diagram of the existing INA control method. The process is as follows: first, the compensation matrix K p (s) is obtained by trial and error, and whether the equivalent object K p (s) G (s) is a diagonal dominant matrix by drawing a Gershgorin band, if not, then Try again to find K p (s); after the equivalent object is a diagonal dominance matrix, set the PID controller and feedback gain and apply it to the CO 2 capture system.
  • FIG. 2 is a schematic block diagram of the improved INA feedforward control method of the present invention.
  • the process is as follows: first, the compensation matrix K p (s) is obtained by the particle swarm optimization algorithm, the diagonally equivalent equivalent object K p (s) G (s) is obtained, then the PID controller is tuned, and finally, the equivalent object is designed before Feed controller, introducing flue gas flow disturbance, outage and CO 2 capture system.
  • FIG. 3 is a principle flowchart of the improved INA method of the present invention. The process is as follows: first, the compensation matrix K p (s) is obtained by trial and error; then the PID controller is set according to the equivalent object of the diagonal advantage; and then the feedback gain is designed according to the inverse equation.
  • Fig. 4-a, Fig. 4-b, Fig. 4-c and Fig. 4-d combined diagrams are schematic diagrams of Gershgorin bands of controlled objects according to the present invention. among them:
  • the abscissa Re represents the real part of the point on the Gershgorin band, and the ordinate lm represents the imaginary part of the point.
  • Figure 4-a is the transfer function of the controlled object Gershgorin band. It can be seen from the figure that the Gershgorin band contains (0,0) points, indicating that it is not a diagonal dominance matrix; Figure 4-b is the transfer function of the controlled object Gershgorin band; Figure 4-c shows the transfer function of the controlled object Gershgorin band; Figure 4-d shows the transfer function of the controlled object The Gershgorin band of Fig. 4-d shows that the Gershgorin band contains (0,0) points, indicating that it is not a diagonal dominant matrix.
  • FIG. 5-a, FIG. 5-b, FIG. 5-c, and FIG. 5-d combined diagrams are schematic diagrams of Gershgorin bands of equivalent objects of the present invention. among them:
  • the abscissa Re represents the real part of the point on the Gershgorin band, and the ordinate lm represents the imaginary part of the point.
  • Figure 5-a is the equivalent object transfer function Gershgorin band. It can be seen from the figure that the Gershgorin band does not contain (0,0) points, indicating that it is a diagonal dominance matrix; Figure 5-b shows the transfer function of the controlled object. Gershgorin band; Figure 5-c shows the transfer function of the controlled object Gershgorin band; Figure 5-d shows the transfer function of the controlled object The Gershgorin band of Fig. 5-d shows that the Gershgorin band does not contain (0,0) points, indicating that it is a diagonally dominant matrix.
  • Figure 6-a, Figure 6-b, Figure 6-c, Figure 6-d and Figure 6-e are the combination of the control method (solid line) and the PI controller (dashed line) without feedforward Schematic comparison of control effects under flow changes. among them:
  • Figure 6-a is the change diagram of the flue gas disturbance
  • Figure 6-b is the change curve of the controlled amount of CO 2 capture rate during the smoke disturbance process under the control method of the present invention and the PI control method without feedforward
  • Fig. 6-c is the control curve of the present invention and the PI control method without feedforward, the change curve of the controlled quantity reboiler temperature during flue gas disturbance
  • Fig. 6-d is the control method of the present invention and no Under the feed-forward PI control method, the change curve of the lean liquid flow during the flue gas disturbance
  • Figure 6-e shows the control method of the present invention and the PI control method without the feedforward. Change curve of extraction steam flow;
  • the control objective is to satisfy the control amount Restricted and rate-constrained, the reboiler temperature was maintained at 386K, and the CO 2 capture rate was tracked to the set value.
  • FIG. 1 is a schematic block diagram of the existing INA control method.
  • the performance parameters and effects of the existing INA control method are detailed in Table 1 above.
  • FIG. 2 discloses a schematic block diagram of the improved INA feedforward control method of the present invention.
  • the specific implementation scheme is:
  • an improved INA feed-forward control method for a post-combustion CO 2 capture system is proposed by the present invention, which is improved by identifying control and controlled quantities of the post-combustion CO 2 capture system.
  • the design of the INA method completes the fusion design of the improved INA and feedforward controller, and then realizes the control of the CO 2 capture rate y 1 and the reboiler temperature y 2.
  • the specific steps include the following:
  • Step 1 Identification of the controlled and controlled quantities of the post-combustion CO 2 capture system: added at steady-state conditions
  • the two control quantities of the lean liquid flow u 1 and the turbine extraction steam flow u 2 and the flue gas disturbance u 3 are the input step excitation signal data, which are sampled every 30s to obtain the CO 2 capture rate y 1 and
  • the two controlled quantities of the reboiler temperature y 2 are the step response data of the output, and the input and output data are identified and used as the controlled open loop transfer function matrix G (s);
  • the flue gas disturbance transfer function is
  • Step 2 Improve the design of the INA method: Use the improved INA method based on particle swarm optimization to decouple the controlled object, obtain the compensation matrix K p (s) through self-optimization, and form an equivalent object G (s) with the controlled object.
  • K p (s); the PID regulator K c (s) diag ⁇ k ci (s) ⁇ is set for the equivalent object, and the setting condition is that the controlled quantity described in step 1 has a smaller Overshoot, fast transition process, closed-loop steady-state error is zero, so as to meet the constraints and rate of change of the control amount described in step 1.
  • Step 3 Improve the integrated design of the INA and the feedforward controller: Based on the decoupled equivalent object, use the invariance principle method to design the feedforward controller to expand and improve the INA to suppress the influence of flue gas disturbances; set the feedforward The feedforward gain of the controller is used to adjust the size of the feedforward effect. If the physical form of the feedforward controller is not achievable, the inertia link is used instead.
  • the inertia link refers to Where T is the inertia time.
  • step 2 of the present invention the particle swarm optimization includes the following steps:
  • the improved INA feedforward control method for a post-combustion CO 2 capture system includes the following specific steps:
  • Step 21 Find the inverse of G (s), draw Gershgorin band, based on the principle that the Gershgorin band does not contain (0,0) points Whether it is a diagonal dominant array;
  • Step 23 Design a PID regulator k ci (s) for the equivalent object G (s) K p (s) to meet the requirements of transient and steady state response;
  • Step 24 Determine the feedback gain f according to the inverse Nai stability criterion, so that the closed-loop system is stable and has an appropriate stability margin;
  • step 25 the post-combustion CO 2 capture system is simulated. If it is not satisfied, return to step 22 to recompensate and readjust k ci (s) until it is satisfied.
  • the particle swarm algorithm described in step 22 includes the following specific steps:
  • the improved INA feedforward control method for a post-combustion CO 2 capture system according to claim 2, wherein the particle swarm algorithm of step 22 includes the following steps:
  • Step 221 the form of the optimized compensation matrix is:
  • the particle swarm is composed of m particles, and each particle has 4 dimensions, which are a, b, c, and d variables, which are variables to be optimized;
  • Step 222 defining a performance function:
  • n is the number of frequency points
  • D point is the distance from the center of the circle to the origin (with the diagonal array g ii (jw) of G (s) as the center), and R is the corresponding radius;
  • the Gershgorin band can be moved away from the origin within the frequency point range, and the optimal value can be obtained within the frequency point range;
  • Step 223, update the particle position:
  • v i (v i1 , v i2 , v i3 , v i4 ) is the velocity of the particle
  • p g (p g1 , p g2 , p g3 , p g4 ) is the optimal position of all particles, upper and lower limits x max , x min , number of particles m, inertia weight ⁇ , history
  • the optimal position learning factor c 1 (x i1 , x i2 , x i3 , x i4 ) represents the current position of each particle
  • p i (p i1 ,
  • Step 224 draw a Gershgorin band to verify whether G (s) K p (s) satisfies the diagonal advantage. If not, change the learning factor c 3 , the frequency point n, and the inertia weight ⁇ , and return to (222) for calculation.
  • the design of the feedforward controller using the principle of invariance described in step 3 refers to: adding the feedforward controller after the PID controller or for the constraints of the output of the feedforward controller, then adding the feedforward controller to the set value ; Then the original coupled object is decomposed into two independent objects by improving the INA method.
  • the transfer function equation in the feedforward controller is:
  • Y (s) represents the output, there are two components of CO 2 capture rate and reboiler temperature, X (s) represents the amount of flue gas input, i takes 1,2 to represent the decoupled
  • W o is the equivalent after compensation
  • W F is the feedforward controller
  • G di is the disturbance transfer function
  • f i is the loop gain
  • M (s), N (s) are the numerator and denominator of the normal physical achievable link, the order of M (s) is lower than N (s), and a is the coefficient of the differential link;
  • the derivation link is physically unachievable, it is transformed into a first-order inertial link, which is equivalent at low frequencies:
  • T is a time constant
  • CO 2 capture rate The set value is stable at 80%, the set value of the reboiler temperature remains unchanged at 386K, and a slope flue gas flow disturbance is added at 500s (that is, from 0.13kg / s to 0.18kg / s).
  • the improved INA feedforward control method of the present invention can eliminate the disturbance Effect, quickly maintain the CO 2 capture rate and reboiler temperature at the set values, compared with PID controller without feedforward, it has a faster and smoother disturbance suppression effect, further improving the CO 2 capture system. Operation quality and adaptability to variable load operation of thermal power plants.
  • the embodiment of the present invention uses the improved INA method of particle swarm optimization to decouple the CO 2 capture system, and on this basis, a feedforward controller is added to suppress the flue gas flow disturbance.
  • a feedforward controller is added to suppress the flue gas flow disturbance.
  • no disturbance Compared with the existing PID method, it has better set value tracking and adjustment capabilities, which improves the control quality of the CO 2 capture system.
  • smoke disturbance In the case of smoke disturbance, it can actively suppress the smoke disturbance and enhance its effectiveness. Adaptability of thermal power plant load changes.
  • the present invention has been verified through repeated tests and has achieved satisfactory trial results.

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)
  • Treating Waste Gases (AREA)

Abstract

本发明涉及一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，它通过对燃烧后CO 2捕集系统的控制量和被控量的辨识，进行改进INA方法的设计，完成改进INA与前馈控制器的融合设计，进而实现对CO 2捕集率y 1和再沸器温度y 2的控制。本发明用于燃烧后CO 2捕集系统后，能够快速平稳追踪变化的CO 2捕集率设定值，有效抑制烟气扰动的影响，具有良好的抗干扰能力。

Description

一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法 技术领域
本发明属于热工控制技术领域，特别是涉及一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法。
背景技术
随着全球变暖日益严重，控制CO 2的产生逐渐成为工业化和城市化进程中的重要任务之一,而我国40％-50％的CO 2来自于火电机组。因此，为了减小燃煤电厂的CO 2排放量，需要在燃煤电厂中增设燃烧后CO 2捕集系统，以此吸收发电过程中产生的CO 2。
燃烧后CO 2捕集系统基于化学吸收法，直接从电厂燃烧后的烟气中分离C0 2,是当前CO 2捕集电站采用的主流技术近年来。而灵活可控性是CO 2捕集系统设计和控制的要求之一。在目前的研究中，最广泛的控制方法是PID调节器控制，如牛红伟等发表的“燃煤电厂烟气CO 2捕集系统的控制策略”一文中，提出利用PID对再生塔溶液循环流量、吸收塔液位等进行协调控制，以达到安全和高效的目的；LinY J等发表的“通过使用单乙醇胺溶液的吸收和抽除来广泛控制CO 2的捕获(Plant wide control of CO 2capture by absorption and stripping using monoethanolamine solution)”一文中，提出了一种分散式PI控制方案，通过控制再沸器温度和贫液流量来维持CO 2捕集率和再沸器温度。这些控制方案虽然能够在无扰动情况下达到基本控制要求，但由于CO 2的吸附和解析涉及到多种设备和化学反应过程,因此系统各变量间存在强耦合和很大的惯性,且系统存在烟气流量等扰动，不考虑解耦或前馈的PID控制方法难以取得满意的控制效果，而目前同时考虑解耦和抑制烟气扰动的燃烧后CO 2捕集系统的控制方法一般采用预测控制方法，但此类方法工程上较难实现。
中国专利申请201710795146.3公开了“一种燃烧后CO 2捕集系统的多模型预测控制方法”，该预测控制方法以基于化学吸附的燃烧后CO 2捕集系统为被控对象，贫液阀门开度和汽轮机低压缸抽汽阀门开度为系统控制输入量，捕集率CO 2和再沸器温度为系统输出量；首先基于子空间辨识方法，利用系统运行产生的数据，在不同工况点处建立系统的局部状态空间模型；接着使用间隙度量的方法调研被控对象的非线性分布；进而在合适的局部工况点处建立预测控制器，并设计隶属度函数将其加权组合，建立燃烧后CO 2捕集系统多模型预测控制系统。该方法虽然具有良好的全局非线性控制能力，能够有效适应系统大范围变工况的需求，快速追踪CO 2捕集率设定值，提高CO 2捕集系统深度快速灵活运行的水平，但还存在明显不足：一是CO 2捕集系统的非线性程度并不明显，采用多模型的方式，并应用了多个预测控制器，使得操作复杂，计算量大，不利于工程实施；二是在工业控制中，通常应用PID方法求解，采用隶属度加权的预测控制方法在工程上难以实现，在变工况时容易出现跳变的现象。
中国专利申请201711138022.4公开了“一种水煤浆气化过程的代理模型建模方法”，该方法选取若干个可测的过程状态作为输入变量，包括氧煤比、煤浆浓度、煤浆流量、煤中H/C元素摩尔比、煤中O/C元素摩尔比及煤中的灰分含量，同时，选取若干个可测的过程状态作为目标输出变量，包括出口合成气中CO的含量、CO 2的含量、H 2的含量、出口处的温度和煤炭中碳的转化率；利用拉丁超立方采样方法对输入变量进行采样后，对输入数据进行分析与处理；利用代理模型Kriging建立输入变量与输出变量之间的数据模型，通过改进后的粒子群优化算法求解出最优代理模型参数。虽然该模型的拟合精度高、跟踪效果好、模型泛化能力强，但还存在明显不足：一是采用的改进粒子群算法仅局限于辨识模型参数，无法应用于燃烧后CO 2捕集系统；二是改进的粒子群算法需要事先进行数据预处理，并应用于离线计算，工业中在扰动情况下若对象发生变化，无法实时获得对象模型。
综上所述，如何克服现有技术所存在的不足已成为当今热工控制技术领域中亟待解决的重点难题之一。
发明内容
本发明的目的是为克服现有技术所存在的不足而提供一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法。本发明用于燃烧后CO 2捕集系统后，能够快速平稳追踪变化的CO 2捕集率设定值，有效抑制烟气扰动的影响，具有良好的抗干扰能力。
根据本发明提出的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，通过对燃烧后CO 2捕集系统的控制量和被控量的辨识，进行改进INA方法的设计，完成改进INA与前馈控制器的融合设计，进而实现对CO 2捕集率y 1和再沸器温度y 2的控制，具体步骤包括如下：
步骤1，对燃烧后CO 2捕集系统的控制量和被控量的辨识：加入在稳态工况下的贫液流量u 1和汽轮机抽汽流量u 2的两个控制量以及烟气扰动量u 3为输入的阶跃激励信号数据，以获取CO 2捕集率y 1和再沸器温度y 2的两个被控量为输出的阶跃响应数据，辨识所述输入输出数据并作为被控开环传递函数阵G(s)；
步骤2，改进INA方法的设计：利用基于粒子群优化的改进INA方法对被控对象解耦，通过自寻优获得补偿矩阵K p(s)，与被控对象组成等效对象G(s)K p(s)；针对所述等效对象整定PID调节器K c(s)＝diag{k ci(s)},所述整定的条件是使步骤1所述的被控量具有较小的超调量、较快的过渡过程、闭环稳态误差为零，以满足步骤1所述的控制量的约束和变化速率；
步骤3，改进INA与前馈控制器的融合设计：基于解耦后的等效对象，采用不变性原理方法设计前馈控制器以拓展改进INA，用于抑制烟气扰动的影响；设前馈控制器的前馈增益以调节前馈作用的大小，若前馈控制器的物理形式不可实现，则采用惯性环节来代替，所述惯性环节是指
Figure PCTCN2019079091-appb-000001
其中T为惯性时间。
本发明提出的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法的进一步优选方案是：
步骤2所述基于粒子群优化的改进INA方法，包括如下具体步骤：
步骤21：求出G(s)的逆，
Figure PCTCN2019079091-appb-000002
绘制
Figure PCTCN2019079091-appb-000003
的Gershgorin带,以Gershgorin带不包含(0,0)点为原则来判断
Figure PCTCN2019079091-appb-000004
是否为对角优势阵；
步骤22，当
Figure PCTCN2019079091-appb-000005
为非对角优势时，利用粒子群算法设计补偿器K p(s)，使
Figure PCTCN2019079091-appb-000006
成为对角优势阵，绘制其Gershgorin带加以验证；
步骤23，对等效对象G(s)K p(s)设计PID调节器k ci(s),以满足暂态与稳态响应的要求；
步骤24，根据逆乃式稳定判据确定反馈增益f，使闭环系统稳定并具有适当的稳定裕度；
步骤25，对燃烧后CO 2捕集系统进行仿真，若不满意，回到步骤22重新补偿，重新调整k ci(s)，直到满意为止。
步骤22所述的粒子群算法，包括如下具体步骤：
步骤221，寻优的补偿矩阵的形式为：
Figure PCTCN2019079091-appb-000007
粒子群由m个粒子组成，每个粒子的维数有4维，分别是a,b,c,d四个变量，作为待寻优的变量；
步骤222，定义性能函数：
定义的性能函数应使找到的
Figure PCTCN2019079091-appb-000008
的Gershgorin带不包含原点，即需要Gershgorin圆心到原点的距离大于半径,并且应保证在给定频率w(s＝jw)中均不包含原点，因此定义的性能函数为：
Figure PCTCN2019079091-appb-000009
(2)式中：n为频率点数，D point是圆心到原点的距离(以G(s)的对角阵g ii(jw)为圆心)，R为对应半径；
根据(2)式，能够在频率点范围内使Gershgorin带远离原点，在频率点范围内得到最优值；
在一段频率点内，根据(2)式计算每个粒子的性能函数，比较每个粒子的性能优劣，根据当前粒子优劣值和历史最优位置进行比对，来调整粒子运动的方向和速度；
步骤223，更新粒子位置：
当获得性能更好的粒子，则替换粒子的最优位置，同时更新各粒子的速度和位置，速度和位置的计算公式如(3)式所示，加入惯性系数ω用于平衡全局搜索能力和局部搜索能力，加入rand()是为防止陷入局部最优，
Figure PCTCN2019079091-appb-000010
使所有粒子都趋向于最优的粒子，当达到最大迭代次数或者全局最优位置满足最小界限时便得到了最终的解；(3)式中：x i＝(x i1,x i2,x i3,x i4)表示每个粒子的当前位置，p i＝(p i1,p i2,p i3,p i4)表示每个粒子的历史最优位置，v i＝(v i1,v i2,v i3,v i4)为粒子的速度，p g＝(p g1,p g2,p g3,p g4)为所有粒子的最优位置，上下限x max，x min，粒子数m，惯性权重ω，历史最优位置的学习因子c 1,全局最优因子c 2和随机学习因子c 3；在设定的解空间内随机初始化每个粒子的位置和速度；
步骤224，绘制Gershgorin带验证G(s)K p(s)是否满足对角优势，若不满足，则更改学习因子c 3,频率点n和惯性权重ω，重新返回(222)进行计算。
步骤(3)所述采用不变性原理设计前馈控制器是指：前馈控制器加在PID控制器之后或者针对需处理前馈控制器输出的约束，则将前馈控制器加在设定值上；然后通过改进INA方法将原本耦合的对象分解为两个独立对象，则前馈控制器中的传递函数方程式为：
Figure PCTCN2019079091-appb-000011
(4)式中：Y(s)表示输出量，有CO 2捕集率和再沸器温度两个分量，X(s)表示烟气输入量，i取1,2分别代表解耦后的两个回路，W o是补偿后的等效对象，W F是前馈控制器,G di表示扰动传递 函数，f i表示回路增益；
若前馈控制器的物理形式不可实现，其传递函数可分解为微分环节和正常的物理可实现环节：
Figure PCTCN2019079091-appb-000012
(5)式中：M(s),N(s)分别是正常物理可实现环节的分子分母，M(s)的阶次低于N(s),a为微分环节的系数；
由于微分环节物理不可实现，将其转化为一阶惯性环节，在低频可等效：
Figure PCTCN2019079091-appb-000013
(6)式中：T为时间常数。
本发明的实现原理是：本发明所涉及的逆奈奎斯特(Inverse Nyquist Array,简称INA)是一种线性多变量系统的控制方法，虽然INA控制方法比现有PID方法在热能控制工程中具有优点，但还没有达到控制的最佳状态，因此本发明提出的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，它是以燃烧后CO 2捕集系统为被控对象、贫液流量和汽轮机抽汽流量为控制量、CO 2捕集率和再沸器温度为被控量以及电站功率变化导致的烟气流量改变为外扰量；首先利用基于粒子群自寻优的改进INA方法对CO 2捕集系统进行解耦，然后针对解耦等效对象利用不变性原理方法引入多变量前馈控制，从而抑制烟气流量扰动的影响，避免了现有INA方法需要反复试凑的缺点。本发明的改进INA前馈控制方法用于燃烧后CO 2捕集系统后，能够快速平稳追踪变化的CO 2捕集率设定值，并有效抑制烟气扰动的影响，具有良好的抗干扰能力。
本发明与现有技术相比其显著优点在于：
第一，本发明具有良好的抗烟气流量扰动干扰能力，并能快速灵活跟踪CO 2捕集率设定值和保持再沸器温度在设定值附近。
第二，本发明相较于现有PID方法，能够更有效的抑制烟气扰动的影响，保证燃烧后CO 2捕集系统灵活运行；在工程上易于实现，特别是在变工况时能够保持平稳运行。
第三，本发明利用基于粒子群优化的改进INA方法对被控对象解耦，使其具有自寻优能力，解决了现有INA需要试凑且效果不显著的缺点。
第四，本发明将改进INA方法和前馈控制器结合，能够实现对可测扰动的抑制，使燃烧后CO 2捕集系统能够快速平稳追踪变化的CO 2捕集率设定值，并有效抑制烟气扰动，具有良好的抗干扰能力。
在烟气量扰动情况下，将改进INA前馈控制方法、现有PID方法和现有INA方法分别控制燃烧后CO 2捕集系统，得到表1所示的控制效果性能参数表。
表1：本发明与现有控制方法的性能参数及效果比较表
Figure PCTCN2019079091-appb-000014
从表1可知，相比于现有PID方法和现有INA方法，改进INA前馈控制方法的调节过渡过 程时间较短，超调量较小，基本没有稳态偏差，说明本发明的控制方法应用于燃烧后CO 2捕集系统能够得到很好的控制效果。
附图说明
图1为现有INA控制方法的原理方框示意图。其流程为：首先通过试凑的方法得到补偿矩阵K p(s)，通过画Gershgorin带的方法来判断等效对象K p(s)G(s)是否为对角优势阵,若不是，则重新试凑K p(s)；在等效对象是对角优势阵后，整定PID控制器和反馈增益，应用于CO 2捕集系统。
图2为本发明的改进INA前馈控制方法的原理方框示意图。其流程为：首先通过粒子群优化算法得到补偿矩阵K p(s)，获得对角优势的等效对象K p(s)G(s)，然后整定PID控制器，最后根据等效对象设计前馈控制器，引入烟气流量扰动，停运与CO 2捕集系统。
图3为本发明的改进INA方法的原理流程图。其流程为：首先通过试凑的方法得到补偿矩阵K p(s)；然后根据对角优势的等效对象，整定PID控制器；之后根据逆乃式判据设计反馈增益。
图4-a、图4-b、图4-c和图4-d组合图为本发明的被控对象的Gershgorin带的示意图。其中：
横坐标Re表示Gershgorin带上点的实部，纵坐标lm表示点的虚部。
图4-a为被控对象传递函数
Figure PCTCN2019079091-appb-000015
的Gershgorin带，从图中可看出Gershgorin带包含(0,0)点，说明其不是对角优势阵；图4-b为被控对象传递函数
Figure PCTCN2019079091-appb-000016
的Gershgorin带；图4-c为被控对象传递函数
Figure PCTCN2019079091-appb-000017
的Gershgorin带；图4-d为被控对象传递函数
Figure PCTCN2019079091-appb-000018
的Gershgorin带，从图4-d中可看出Gershgorin带包含(0,0)点，说明其不是对角优势阵。
图5-a、图5-b、图5-c和图5-d组合图为本发明的等效对象的Gershgorin带的示意图。其中：
横坐标Re表示Gershgorin带上点的实部，纵坐标lm表示点的虚部。
图5-a为等效对象传递函数
Figure PCTCN2019079091-appb-000019
的Gershgorin带，从图中可看出Gershgorin带不包含(0,0)点，说明其是对角优势阵；图5-b为被控对象传递函数
Figure PCTCN2019079091-appb-000020
的Gershgorin带；图5-c为被控对象传递函数
Figure PCTCN2019079091-appb-000021
的Gershgorin带；图5-d为被控对象传递函数
Figure PCTCN2019079091-appb-000022
的Gershgorin带，从图5-d中可看出Gershgorin带不包含(0,0)点，说明其是对角优势阵。
图6-a、图6-b、图6-c、图6-d和图6-e组合图为本发明控制方法(实线)与不加前馈的PI控制器(虚线)在烟气流量变化下的控制效果对比的示意图。其中：
图6-a为烟气扰动的变化图；图6-b为本发明的控制方法和不加前馈的PI控制方法下，烟气扰动过程中被控量CO 2捕集率的变化曲线；图6-c为本发明的控制方法和不加前馈的PI控制方 法下，烟气扰动过程中被控量再沸器温度的变化曲线；图6-d为本发明的控制方法和不加前馈的PI控制方法下，烟气扰动过程中控制量贫液流量的变化曲线；图6-e为本发明的控制方法和不加前馈的PI控制方法下，烟气扰动过程中控制量抽汽流量的变化曲线；
具体实施方式
下面结合附图和实施例对本发明的具体实施方案做进一步的详细描述。
以本发明提出的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法在某1MW火电机组燃烧后CO 2捕集系统仿真模型中应用为例，控制目标是在满足控制量大小约束和速率约束的调节下，使再沸器温度保持386K，使CO 2捕集率跟踪设定值。
图1为现有INA控制方法的原理方框示意图，应用现有INA控制方法的性能参数及效果详见上述表1。为克服现有INA控制方法的不足，图2公开了本发明的改进INA前馈控制方法的原理方框示意图，具体实施方案是：
如图2所示，本发明提出的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，通过对燃烧后CO 2捕集系统的控制量和被控量的辨识，进行改进INA方法的设计，完成改进INA与前馈控制器的融合设计，进而实现对CO 2捕集率y 1和再沸器温度y 2的控制，具体步骤包括如下：
步骤1，对燃烧后CO 2捕集系统的控制量和被控量的辨识：加入在稳态工况
Figure PCTCN2019079091-appb-000023
下的贫液流量u 1和汽轮机抽汽流量u 2的两个控制量以及烟气扰动量u 3为输入的阶跃激励信号数据，每30s采样一次，以获取CO 2捕集率y 1和再沸器温度y 2的两个被控量为输出的阶跃响应数据，辨识所述输入输出数据并作为被控开环传递函数阵G(s)；
Figure PCTCN2019079091-appb-000024
烟气扰动传递函数为
Figure PCTCN2019079091-appb-000025
步骤2，改进INA方法的设计：利用基于粒子群优化的改进INA方法对被控对象解耦，通过自寻优获得补偿矩阵K p(s)，与被控对象组成等效对象G(s)K p(s)；针对所述等效对象整定PID调节器K c(s)＝diag{k ci(s)}，所述整定的条件是使步骤1所述的被控量具有较小的超调量、较快 的过渡过程、闭环稳态误差为零，以满足步骤1所述的控制量的约束和变化速率；
Figure PCTCN2019079091-appb-000026
步骤3，改进INA与前馈控制器的融合设计：基于解耦后的等效对象，采用不变性原理方法设计前馈控制器以拓展改进INA，用于抑制烟气扰动的影响；设前馈控制器的前馈增益以调节前馈作用的大小，若前馈控制器的物理形式不可实现，则采用惯性环节来代替，所述惯性环节是指
Figure PCTCN2019079091-appb-000027
其中T为惯性时间。
如图3所示，在上述本发明步骤2中，基于粒子群优化包括如下步骤：
根据权利要求1所述的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，其特征在于：步骤2所述基于粒子群优化的改进INA方法，包括如下具体步骤：
步骤21：求出G(s)的逆，
Figure PCTCN2019079091-appb-000028
绘制
Figure PCTCN2019079091-appb-000029
的Gershgorin带,以Gershgorin带不包含(0,0)点为原则来判断
Figure PCTCN2019079091-appb-000030
是否为对角优势阵；
如图4-a、图4-b、图4-c和图4-d的组合图所示，具体为
Figure PCTCN2019079091-appb-000031
的Gershgorin带，从该组合图中可看出
Figure PCTCN2019079091-appb-000032
包含原点，说明
Figure PCTCN2019079091-appb-000033
不是对角优势阵，因此需要获取补偿阵
Figure PCTCN2019079091-appb-000034
步骤22，当
Figure PCTCN2019079091-appb-000035
为非对角优势时，利用粒子群算法设计补偿器K p(s)，使
Figure PCTCN2019079091-appb-000036
成为对角优势阵，绘制其Gershgorin带加以验证；
如图5-a、图5-b、图5-c和图5-d的组合图所示，具体为
Figure PCTCN2019079091-appb-000037
的Gershgorin带，从该组合图中可看出
Figure PCTCN2019079091-appb-000038
包含原点，是对角优势阵；
步骤23，对等效对象G(s)K p(s)设计PID调节器k ci(s),以满足暂态与稳态响应的要求；
步骤24，根据逆乃式稳定判据确定反馈增益f，使闭环系统稳定并具有适当的稳定裕度；
步骤25，对燃烧后CO 2捕集系统进行仿真，若不满意，回到步骤22重新补偿，重新调整k ci(s)，直到满意为止。
步骤22，当
Figure PCTCN2019079091-appb-000039
为非对角优势时，利用粒子群算法设计补偿器K p(s)，使
Figure PCTCN2019079091-appb-000040
成为对角优势阵，绘制其Gershgorin带加以验证；
其中，步骤22步骤所述粒子群算法，包括如下具体步骤：
根据权利要求2所述的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，其特征在于：步骤22所述的粒子群算法，包括如下步骤：
步骤221，寻优的补偿矩阵的形式为：
Figure PCTCN2019079091-appb-000041
粒子群由m个粒子组成，每个粒子的维数有4维，分别是a,b,c,d四个变量，作为待寻优的变量；
步骤222，定义性能函数：
定义的性能函数应使找到的
Figure PCTCN2019079091-appb-000042
的Gershgorin带不包含原点，即需要Gershgorin圆心到原点的距离大于半径,并且应保证在给定频率w(s＝jw)中均不包含原点，因此定义的性能函数为：
Figure PCTCN2019079091-appb-000043
(3)式中：n为频率点数，D point是圆心到原点的距离(以G(s)的对角阵g ii(jw)为圆心)，R为对应半径；
根据(5)式，能够在频率点范围内使Gershgorin带远离原点，在频率点范围内得到最优值；
在一段频率点内，根据(5)式计算每个粒子的性能函数，比较每个粒子的性能优劣，根据当前粒子优劣值和历史最优位置进行比对，来调整粒子运动的方向和速度；
步骤223，更新粒子位置：
当获得性能更好的粒子，则替换粒子的最优位置，同时更新各粒子的速度和位置，速度和位置的计算公式如(6)式所示，加入惯性系数ω用于平衡全局搜索能力和局部搜索能力，加入rand()是为防止陷入局部最优，
Figure PCTCN2019079091-appb-000044
使所有粒子都趋向于最优的粒子，当达到最大迭代次数或者全局最优位置满足最小界限时便得到了最终的解；(6)式中：x i＝(x i1,x i2,x i3,x i4)表示每个粒子的当前位置，p i＝(p i1,p i2,p i3,p i4)表示每个粒子的历史最优位置，v i＝(v i1,v i2,v i3,v i4)为粒子的速度，p g＝(p g1,p g2,p g3,p g4)为所有粒子的最优位置，上下限x max，x min，粒子数m，惯性权重ω，历史最优位置的学习因子c 1,全局最优因子c 2和随机学习因子c 3；在设定的解空间内随机初始化每个粒子的位置和速度；
步骤224，绘制Gershgorin带验证G(s)K p(s)是否满足对角优势，若不满足，则更改学习因子c 3,频率点n和惯性权重ω，重新返回(222)进行计算。
使所有粒子都趋向于最优的粒子，当达到最大迭代次数或者全局最优位置满足最小界限时便得到了最终的解。最终找到的补偿矩阵为
Figure PCTCN2019079091-appb-000045
步骤3所述采用不变性原理设计前馈控制器是指：前馈控制器加在PID控制器之后或者针对需处理前馈控制器输出的约束，则将前馈控制器加在设定值上；然后通过改进INA方法将原本耦合的对象分解为两个独立对象，则前馈控制器中的传递函数方程式为：
Figure PCTCN2019079091-appb-000046
(8)式中：Y(s)表示输出量，有CO 2捕集率和再沸器温度两个分量，X(s)表示烟气输入量，i取1,2分别代表解耦后的两个回路，W o是补偿后的等效对象，W F是前馈控制器,G di表示扰动传递函数，f i表示回路增益；
若前馈控制器的物理形式不可实现，其传递函数可分解为微分环节和正常的物理可实现环节：
Figure PCTCN2019079091-appb-000047
(9)式中：M(s),N(s)分别是正常物理可实现环节的分子分母，M(s)的阶次低于N(s),a为微分环节的系数；
由于微分环节物理不可实现，将其转化为一阶惯性环节，在低频可等效：
Figure PCTCN2019079091-appb-000048
(10)式中：T为时间常数。
本实施例为了比较本发明中的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法和不带前馈的PID控制方法的控制效果，做如下仿真试验：CO 2捕集率设定值稳定在80％，再沸器温度设定值保持386K不变，在500s加入一个斜坡烟气流量扰动(即：从0.13kg/s变化到0.18kg/s)。
如图6-a、图6-b、图6-c、图6-d和图6-e的组合图所示，当烟气扰动发生时，本发明的改进INA前馈控制方法能够消除扰动的影响，快速维持CO 2捕集率和再沸器温度在设定值上，相较于不带前馈的PID控制器，具有更快速平稳的扰动抑制效果，进一步提高CO 2捕集系统的运行品质以及对火电厂变负荷运行的适应性。
综上所述，本发明实施例运用了粒子群优化的改进INA方法对CO 2捕集系统进行解耦，在此基础上加入前馈控制器抑制烟气流量扰动，在无扰的情况下，与现有PID方法相比具有更优的设定值跟踪和调节能力，提高了CO 2捕集系统的控制品质；在有烟气扰动的情况下，可主动抑制烟气扰动，增强了其对火电厂负荷变动的适应性。
本发明的具体实施方式中凡未涉及的说明属于本领域的公知技术，可参考公知技术加以实施。
本发明经反复试验验证，取得了满意的试用效果。
以上具体实施方式及实施例是对本发明提出的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法技术思想的具体支持，不能以此限定本发明的保护范围，凡是按照本发明提出的技术思想，在本技术方案基础上所做的任何等同变化或等效的改动，均仍属于本发明技术方案保护的范围。

Claims (4)

  1. 一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，其特征在于，
    通过对燃烧后CO 2捕集系统的控制量和被控量的辨识，进行改进INA方法的设计，完成改进INA与前馈控制器的融合设计，进而实现对CO 2捕集率y 1和再沸器温度y 2的控制，具体步骤包括如下：
    步骤1，对燃烧后CO 2捕集系统的控制量和被控量的辨识：加入在稳态工况下的贫液流量u 1和汽轮机抽汽流量u 2的两个控制量以及烟气扰动量u 3为输入的阶跃激励信号数据，以获取CO 2捕集率y 1和再沸器温度y 2的两个被控量为输出的阶跃响应数据，辨识所述输入输出数据并作为被控开环传递函数阵G(s)；
    步骤2，改进INA方法的设计：利用基于粒子群优化的改进INA方法对被控对象解耦，通过自寻优获得补偿矩阵K p(s)，与被控对象组成等效对象G(s)K p(s)；针对所述等效对象整定PID调节器K c(s)＝diag{k ci(s)}，所述整定的条件是使步骤1所述的被控量具有较小的超调量、较快的过渡过程、闭环稳态误差为零，以满足步骤1所述的控制量的约束和变化速率；
    步骤3，改进INA与前馈控制器的融合设计：基于解耦后的等效对象，采用不变性原理方法设计前馈控制器以拓展改进INA，用于抑制烟气扰动的影响；设前馈控制器的前馈增益以调节前馈作用的大小，若前馈控制器的物理形式不可实现，则采用惯性环节来代替，所述惯性环节是指
    Figure PCTCN2019079091-appb-100001
    其中T为惯性时间。
  2. 根据权利要求1所述的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，其特征在于：步骤2所述基于粒子群优化的改进INA方法，包括如下具体步骤：
    步骤21：求出G(s)的逆，
    Figure PCTCN2019079091-appb-100002
    绘制
    Figure PCTCN2019079091-appb-100003
    的Gershgorin带,以Gershgorin带不包含(0,0)点为原则来判断
    Figure PCTCN2019079091-appb-100004
    是否为对角优势阵；
    步骤22，当
    Figure PCTCN2019079091-appb-100005
    为非对角优势时，利用粒子群算法设计补偿器K p(s)，使
    Figure PCTCN2019079091-appb-100006
    成为对角优势阵，绘制其Gershgorin带加以验证；
    步骤23，对等效对象G(s)K p(s)设计PID调节器k ci(s),以满足暂态与稳态响应的要求；
    步骤24，根据逆乃式稳定判据确定反馈增益f，使闭环系统稳定并具有适当的稳定裕度；
    步骤25，对燃烧后CO 2捕集系统进行仿真，若不满意，回到步骤22重新补偿，重新调整k ci(s)，直到满意为止。
  3. 根据权利要求2所述的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，其特征在于：步骤22所述的粒子群算法，包括如下步骤：
    步骤221，寻优的补偿矩阵的形式为：
    Figure PCTCN2019079091-appb-100007
    粒子群由m个粒子组成，每个粒子的维数有4维，分别是a,b,c,d四个变量，作为待寻优的变量；
    步骤222，定义性能函数：
    定义的性能函数应使找到的
    Figure PCTCN2019079091-appb-100008
    的Gershgorin带不包含原点，即需要Gershgorin圆心到原点的距离大于半径,并且应保证在给定频率w(s＝jw)中均不包含原点，因此定义的性能函数为：
    Figure PCTCN2019079091-appb-100009
    (2)式中：n为频率点数，D point是圆心到原点的距离(以G(s)的对角阵g ii(jw)为圆心)，R为对应半径；
    根据(2)式，能够在频率点范围内使Gershgorin带远离原点，在频率点范围内得到最优值；
    在一段频率点内，根据(2)式(2)计算每个粒子的性能函数，比较每个粒子的性能优劣，根据当前粒子优劣值和历史最优位置进行比对，来调整粒子运动的方向和速度；
    步骤223，更新粒子位置：
    当获得性能更好的粒子，则替换粒子的最优位置，同时更新各粒子的速度和位置，速度和位置的计算公式如(3)式所示，加入惯性系数ω用于平衡全局搜索能力和局部搜索能力，加入rand()是为防止陷入局部最优，
    Figure PCTCN2019079091-appb-100010
    使所有粒子都趋向于最优的粒子，当达到最大迭代次数或者全局最优位置满足最小界限时便得到了最终的解；(3)式中：x i＝(x i1,x i2,x i3,x i4)表示每个粒子的当前位置，p i＝(p i1,p i2,p i3,p i4)表示每个粒子的历史最优位置，v i＝(v i1,v i2,v i3,v i4)为粒子的速度，p g＝(p g1,p g2,p g3,p g4)为所有粒子 的最优位置，上下限x max，x min，粒子数m，惯性权重ω，历史最优位置的学习因子c 1,全局最优因子c 2和随机学习因子c 3；在设定的解空间内随机初始化每个粒子的位置和速度；
    步骤224，绘制Gershgorin带验证G(s)K p(s)是否满足对角优势，若不满足，则更改学习因子c 3,频率点n和惯性权重ω，重新返回(222)进行计算。
  4. 根据权利要求1所述的一种用于燃烧后CO 2捕集系统的改进INA前馈控制方法，其特征在于：步骤(3)所述采用不变性原理设计前馈控制器是指：前馈控制器加在PID控制器之后或者针对需处理前馈控制器输出的约束，则将前馈控制器加在设定值上；然后通过改进INA方法将原本耦合的对象分解为两个独立对象，则前馈控制器中的传递函数方程式为：
    Figure PCTCN2019079091-appb-100011
    (4)式中：Y(s)表示输出量，有CO 2捕集率和再沸器温度两个分量，X(s)表示烟气输入量，i取1,2分别代表解耦后的两个回路，W o是补偿后的等效对象，W F是前馈控制器,G di表示扰动传递函数，f i表示回路增益；
    若前馈控制器的物理形式不可实现，其传递函数可分解为微分环节和正常的物理可实现环节：
    Figure PCTCN2019079091-appb-100012
    (5)式中：M(s),N(s)分别是正常物理可实现环节的分子分母，M(s)的阶次低于N(s),a为微分环节的系数；
    由于微分环节物理不可实现，将其转化为一阶惯性环节，在低频可等效：
    Figure PCTCN2019079091-appb-100013
    (6)式中：T为时间。
PCT/CN2019/079091 2018-09-28 2019-03-21 一种用于燃烧后co 2捕集系统的改进ina前馈控制方法 WO2020062806A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811137761.6A CN109188911B (zh) 2018-09-28 2018-09-28 一种用于燃烧后co2捕集系统的改进ina前馈控制方法
CN201811137761.6 2018-09-28

Publications (1)

Publication Number Publication Date
WO2020062806A1 true WO2020062806A1 (zh) 2020-04-02

Family

ID=64906796

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/079091 WO2020062806A1 (zh) 2018-09-28 2019-03-21 一种用于燃烧后co 2捕集系统的改进ina前馈控制方法

Country Status (2)

Country Link
CN (1) CN109188911B (zh)
WO (1) WO2020062806A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109188911B (zh) * 2018-09-28 2019-07-23 东南大学 一种用于燃烧后co2捕集系统的改进ina前馈控制方法
CN110687792B (zh) * 2019-11-04 2022-04-26 东南大学 一种用于化学吸附燃烧后二氧化碳捕集系统的抗烟气扰动模糊控制方法
CN114397922B (zh) * 2021-09-29 2023-03-14 北京百利时能源技术股份有限公司 一种燃煤电厂二氧化碳捕集再沸器温度控制系统
CN114011865B (zh) * 2021-11-05 2022-09-23 北京航空航天大学 一种应用于土壤气相抽提设备的多变量解耦控制方法
CN114712989B (zh) * 2022-03-11 2023-01-03 浙江大学 低成本高效的污染物与co2协同吸收-解吸解耦方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102502631A (zh) * 2011-09-30 2012-06-20 清华大学 基于不同灵活运行模式的燃烧后碳捕集电厂控制方法
WO2013144888A1 (en) * 2012-03-30 2013-10-03 Alstom Technology Ltd Integrated electric power generation and steam demand control system for a post combustion co2 capture plants
CN106842955A (zh) * 2017-03-15 2017-06-13 东南大学 带烟气量扰动抑制的燃烧后co2捕集系统预测控制方法
CN107844659A (zh) * 2017-11-16 2018-03-27 华东理工大学 水煤浆气化过程的代理模型建模方法
CN109188911A (zh) * 2018-09-28 2019-01-11 东南大学 一种用于燃烧后co2捕集系统的改进ina前馈控制方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7966829B2 (en) * 2006-12-11 2011-06-28 General Electric Company Method and system for reducing CO2 emissions in a combustion stream
CN107450325B (zh) * 2017-09-06 2019-11-12 东南大学 一种燃烧后co2捕集系统的多模型预测控制方法
CN108508748B (zh) * 2018-03-28 2020-11-03 东南大学 一种有效的光热辅助燃烧后co2捕集系统运行控制方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102502631A (zh) * 2011-09-30 2012-06-20 清华大学 基于不同灵活运行模式的燃烧后碳捕集电厂控制方法
WO2013144888A1 (en) * 2012-03-30 2013-10-03 Alstom Technology Ltd Integrated electric power generation and steam demand control system for a post combustion co2 capture plants
CN106842955A (zh) * 2017-03-15 2017-06-13 东南大学 带烟气量扰动抑制的燃烧后co2捕集系统预测控制方法
CN107844659A (zh) * 2017-11-16 2018-03-27 华东理工大学 水煤浆气化过程的代理模型建模方法
CN109188911A (zh) * 2018-09-28 2019-01-11 东南大学 一种用于燃烧后co2捕集系统的改进ina前馈控制方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
KERSHENBAUM, L.S. ET AL.: "Implementation of On-line Control in Chemical Process Plants", AUTOMATICA, vol. 17, no. 6, 31 December 1981 (1981-12-31), pages 777 - 788, XP055699655, ISSN: 0005-1098 *
YANG, MINGXING ET AL.: "Research on INA Method-Based Rolling Missile Decoupling", COMPUTER SIMULATION, vol. 28, no. 5, 31 May 2011 (2011-05-31), ISSN: 1006-9348 *

Also Published As

Publication number Publication date
CN109188911B (zh) 2019-07-23
CN109188911A (zh) 2019-01-11

Similar Documents

Publication Publication Date Title
WO2020062806A1 (zh) 一种用于燃烧后co 2捕集系统的改进ina前馈控制方法
Wu et al. Superheated steam temperature control based on modified active disturbance rejection control
Wu et al. On transitioning from PID to ADRC in thermal power plants
Sun et al. Enhanced decentralized PI control for fluidized bed combustor via advanced disturbance observer
CN105629738B (zh) Scr烟气脱硝系统控制方法及设备
Sun et al. Direct energy balance based active disturbance rejection control for coal-fired power plant
CN108490790A (zh) 一种基于多目标优化的过热汽温自抗扰串级控制方法
Prasad et al. Plant-wide predictive control for a thermal power plant based on a physical plant model
CN104534507A (zh) 一种锅炉燃烧优化控制方法
CN113433911B (zh) 基于浓度精准预测的脱硝装置喷氨精准控制系统与方法
Wei et al. ${H} _ {\infty} $-LQR-Based Coordinated Control for Large Coal-Fired Boiler–Turbine Generation Units
CN102425790B (zh) 一种循环流化床锅炉在线优化自学习控制方法
CN105786035A (zh) 基于试探式预测控制技术的火电机组scr脱硝优化控制系统
Zhang et al. Model predictive control for load frequency control with wind turbines
Qiao A novel online modeling for NOx generation prediction in coal-fired boiler
CN105889910B (zh) 一种循环流化床锅炉的新型agc控制方法
Xu et al. Design of type-2 fuzzy fractional-order proportional-integral-derivative controller and multi-objective parameter optimization under load reduction condition of the pumped storage unit
Tang et al. Predictive control of SCR denitrification system in thermal power plants based on GA‐BP and PSO
Lian Main steam temperature control based on variable universe fuzzy dynamic matrix control
Prasad et al. A hierarchical physical model-based approach to predictive control of a thermal power plant for ef” cient plant-wide disturbance rejection
Li et al. Performance analysis of reheat steam temperature control system of thermal power unit based on constrained predictive control
Zhu et al. Internal model control using LMS filter and its application to superheated steam temperature of power plant
Liu et al. Smooth sliding mode control and its application in ship boiler drum water level
Hu et al. Adaptive internal model control of SCR denitration system based on multi-objective optimization
Wu et al. Frequency-domain analysis of a modified active disturbance rejection control with application to superheated steam temperature control

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19867223

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19867223

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19867223

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205 DATED 13.01.2023)

122 Ep: pct application non-entry in european phase

Ref document number: 19867223

Country of ref document: EP

Kind code of ref document: A1