http://www.paper.edu.

cn
IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 13, NO. 6, NOVEMBER 2005 943

A New Coordinated Control Strategy

for Boiler-Turbine System of
Coal-Fired Power Plant
Shaoyuan Li, Senior Member, IEEE, Hongbo Liu, Wen-Jian Cai, Yeng-Chai Soh, and Li-Hua Xie

Abstract—This paper presents the new development of the Differential parameter in PID controller.
boiler-turbine coordinated control strategy using fuzzy rea- Transfer function output to input .
soning and autotuning techniques. The boiler-turbine system is Decoupling compensator output matrix.
a very complex process that is a multivariable, nonlinear, slowly
time-varying plant with large settling time and a lot of uncer- Decoupling compensator matrix.
tainties. As there exist strong couplings between the main steam Error between set-point and current value.
pressure control loop and the power output control loop in the Change value of the .
boiler-turbine unit with large time-delay and uncertainties, auto- First-order lag filter time constant.
matic coordinated control of the two loops is a very challenging System sampling period.
problem. This paper presents a new coordinated control strategy
(CCS) which is organized into two levels: a basic control level and Fuzzy membership of variable .
a high supervision level. Proportional-integral derivative (PID) System static gain.
type controllers are used in the basic level to perform basic control System critical period.
functions while the decoupling between two control loops can be Amplitude margin.
realized in the high level. A special subclass of fuzzy inference Phase margin.
systems, called the Gaussian partition with evenly (GPE) spaced
midpoints systems, is used to self-tune the main steam pressure Boiler firing rate.
PID controller’s parameters online based on the error signal and Governor value position.
its first difference, aimed at overcoming the uncertainties due to Main steam pressure.
changing fuel calorific value, machine wear, contamination of the Power output.
boiler heating surfaces and plant modeling errors. For the large Main steam flow.
variation of operating condition, a supervisory control level has
been developed by autotuning technique. The developed CCS has Main steam temperature.
been implemented in a power plant in China, and satisfactory Abbreviations
industrial operation results demonstrate that the proposed control DCS Distributed control systems.
strategy has enhanced the adaptability and robustness of the CCS Coordinated control system.
process. Indeed, better control performance and economic benefit GPE Gaussian partition with evenly space.
have been achieved.
TPE Triangle partition with evenly space.
Index Terms—Boiler-turbine coordinated control strategy, CARMA Controlled autoregressive moving average.
decoupling control, industrial application, multivariable systems, AGC Automatic generation control.
power plant.

II. INTRODUCTION
I. NOMENCLATURE
Critical gain in power control loop.
Critical gain in pressure control loop.
T HE majority of coal-fired power plants in China built more
than a decade ago were once expected to operate at near
full capacity but are now operating in a load following mode
Proportional parameter in proportional-integral due to the rapid development of the power industry. Conse-
derivative (PID) controller. quently, AGC of power networks becomes necessary in order
Integral parameter in PID controller. to meet varying load demands at different time periods. As a
key component of power network AGC, the coordinated con-
Manuscript received December 1, 2003. Manuscript received in final form trol of fossil-fueled generating units plays a vital role in safe
June 8, 2005. Recommended by Associate Editor V. Gopal. This work was and economic operation of the system. Since the performance
supported in part by the National Natural Science Foundation of China under
Grant 60474051, in part by the Key Technology and Development Program of of the multi-input–multi-output (MIMO) boiler-turbine system
Shanghai Science and Technology Department under Grant 04DZ11008, and in can vary significantly due to the complex nonlinearity in dif-
part by the program for New Century Excellent Talents in University of China ferent operating regions, conventional linear control methods
(NCET).
S. Li and H. Liu are with the Institute of Automation, Shanghai Jiaotong Uni- may not be sufficient for the whole operation range [1]–[5].
versity, Shanghai 200030, China (e-mail: syli@ sjtu.edu.cn). For the boiler-turbine control system, the central task is to
W.-J. Cai, Y.-C. Soh, and L.-H. Xie are with the School of Electrical and adjust the output power to meet system demand while min-
Electronic Engineering, Nanyang Technological University, Singapore 639798,
Singapore (e-mail: [email protected]). imizing unwanted pressure and temperature variations. The
Digital Object Identifier 10.1109/TCST.2005.854319 turbine speed is controlled by the main steam pressure to drive

1063-6536/$20.00 © 2005 IEEE

中国科技论文在线
944 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 13, NO. 6, NOVEMBER 2005

the power generator [1]. However, as the electricity demand

swings from minute to minute, the turbine speed has to be
varied to meet the different load demands. The variation of
turbine speed causes a chain reaction of change in firing rate,
demand on the coal, grinding and feeding coal, and eventually
the steam generated. The time scale of the steam pressure
process is naturally quite slow at about 8 to 15 min for the
system under study. In contrast, by opening the governor
valve, different amount of steam can be supplied immediately
[2], but this is at the expense of depleting stored energy in
the evaporator of the boiler leading to main steam pressure
variations, and it takes a long time to recover it to its reference
value. Thus, for a rapidly changing power demand, controlling
the governor valve will be more effective, but a sustained
change can only be achieved via changing the firing rate. As
a compromise, a master–slave control strategy [3] has been Fig. 1. Schematic diagram of the 300 MW boiler-turbine unit.
adopted in most power plants in conventional control mode,
i.e., the turbine speed control loop works as the master loop to
track the main steam pressure, and the combustion control loop
as the slave loop to track the varying turbine speed. In such a
scheme, a proportional derivative (PD) controller is needed to
coordinate the relationship between the two control loops [3],
the controller parameters, however, are very difficult to adjust
even within a very narrow operation range. In practice, it is
mostly done by a trial-and-error method. As a result, the main
steam pressure and the power output of the generating units, in
most systems, are still being controlled manually. It is difficult Fig. 2. Conventional CCS.
to avoid excessive stresses on the process components and to
meet the economic and quality requirement.
Even though several advanced control structures for the of the main steam pressure online, and it overcomes
boiler-turbine CCS have been proposed to tackle the problem uncertainties caused by changing fuel calorific value,
in the literature, the problem still remains unsolved for a large machine wear and plant modeling errors.
operating range [6]–[9]. In this paper, an advanced control The control strategy has been realized in the FOXBORO I/A
strategy is proposed to solve this particular coordinated control Series DCS and implemented on a 300 MW boiler-turbine unit,
problem. By considering the complex nature of the process, a i.e., Unit 1 of Yuanbaoshan Power Plant in China for two years.
supervisory control structure is proposed and it consists of the The remaining sections of this paper are organized as follows.
following components. Section II describes the process and plant characteristics. The
steady-state triangular decoupler and autotuning supervisory
1) Proportional–integral derivative (PID)-type controllers control techniques are given in Section III. Section IV discusses
are used as basic control units due to their simple struc- the gain-scheduling of the PID type controllers and decouplers
ture and the concept is well understood by field engineers and the autotuning main steam pressure PID controller by
and operators. a fuzzy mechanism. Section V presents the implementation
2) A steady-state triangular decoupler is designed which si- and the operating results of the proposed control strategy on
multaneously decouples the strongly coupled main steam the boiler-turbine system. Finally, conclusions are drawn in
pressure and power output loops for both set-point and un- Section VI.
measured pulverized coal disturbance.
3) A supervisory control level via an autotuning technique
is used to tune the parameters of the controller at dif- III. PROCESS DESCRIPTION
ferent operating conditions as system parameters can vary The schematic diagram of the investigated 300 MW boiler-
significantly due to the complex nonlinearity in load fol- turbine unit is shown in Fig. 1. The boiler is a compound cir-
lowing mode. culation tower boiler with low circulation ratio that produces
4) The PID type controllers and decouplers are gain-sched- 947 tons of steam per hour at maximum continuous rating. The
uled according to the actual load to take account of the rated main steam pressure and the super-heater outlet temper-
load-dependent nonlinear characteristics of the boiler-tur- ature are 18.5 Mpa and 545 C, respectively. Electric power is
bine process. generated by the feeding steam from the boiler to the turbine,
5) A special subclass of fuzzy inference systems, i.e., and the steam from the turbine condenses to water through the
Gaussian partition system with evenly spaced midpoints condenser, which is sent to the boiler again by the feedwater
(GPE) [17], is employed to autotune the PID parameters pump.
LI et al.: A NEW COORDINATED CONTROL STRATEGY FOR BOILER-TURBINE SYSTEM http://www.paper.edu.cn
945

Fig. 3. New CCS.

A characteristic of the boiler system is highly complex and pulverized coal disturbances caused by uncertain coal
nonlinear. The main nonlinearity is related to the property of mill working conditions [3], the strong couplings could
gain and time constant in the boiler process, and the property create severe problems for system stability and control,
variation is dependent on the plant load. In order to control the particularly when the unit is operating at high load.
complex system, the boiler-turbine CCS is usually adopted in 2) Nonlinearity. The system exhibits highly nonlinear char-
the industry which can be regarded as a two-input–two-output acteristics when the power output changes over a wide
(TITO) multivariable control system. The conventional struc- range, as pressure increases/decreases and heat transfer
ture of CCS is shown in Fig. 2, where the two inputs are boiler rates do not vary in proportion to the fuel inputs.
firing rate (coal feeder speed) and turbine governor valve po- 3) Long Settling Time with Time-delay. There is a long
sition ; two outputs are the main steam pressure and the settling time in the change of steam pressure or power
real power output , respectively. output while changing the coal feeder speed due to mas-
The boiler-turbine unit is a time-varying and nonlinear system sive storages of mass and energy. The results of real-time
with strong interactions and uncertainties. Through careful the- dynamic experiments [5] show that the time constant and
oretical analysis [5], a linearized system model for a given op- time-delay from the coal feeder speed to the main steam
erating point can be obtained as pressure or the power output can vary within 6 15 min
with varying load.
4) Uncertainty. In coal-fired units, the fuel is supplied by
coal mills which have uncertain dynamics due to varying
(1) time-delay of the grinding process and the uncertain be-
havior of the mills caused by coal quality variation and
where the and denotes the steady-state values of and machine wear.
, is the transfer function between the main steam flow
and the power output which represents the dynamics of IV. COORDINATED CONTROL
the turbine and reheater.
Although the boiler-turbine unit transfer function can be ap- A hierarchical coordinated control scheme for the main steam
proximated by (1) for a given operating point, the boiler-tur- pressure and the power output consists of two levels: a basic
bine system remains very complex, and conventional control ap- control level and a supervisory control level as shown in Fig. 3.
proaches [1], [5] have encountered great difficulties due to the The supervisory control level is referred to as the autotuner
following factors. of CCS controller/decoupler, its functions include:
1) Strong Coupling. There are strong couplings between • monitoring the control performance online and iden-
the main steam pressure control loop and the power tifying the mathematical model of the boiler-turbine
output control loop. If there exists frequent unmeasured system;
中国科技论文在线
946 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 13, NO. 6, NOVEMBER 2005

Fig. 4. Multivariable feedback structure of CCS.

• retuning CCS controller/decoupler parameters automati-

cally to enhance the adaptability of the control system due
to large variation of operating conditions.
Fig. 5. Z-N identification scheme of CCS.
The basic control level consists of three conventional feedfor-
ward/feedback PID-type controllers:
• A unilateral PD-type decoupler adjusts the power output 1) Closed-loop identification of the controlled plant by using
to follow load demand changes while maintaining the the decentralized step test and least-squares algorithm
main steam pressure within the permitted range and com- around an operating condition based on the input–output
pensates the load varying caused by the uncertain coal plant data (see the Appendix).
mill working condition. 2) Determine the static gain in (2) according to the con-
• A PI controller for the electric power output loop. trolled plant model to adapt the decoupling compensator
• A fuzzy autotuning PID controller in the main steam pres- matrix .
sure control loop which provides good performance with 3) Perform multivariable Ziegler–Nichols identification
the ability to overcome uncertainties due to changing fuel experiments [16] according to Fig. 5 using the dynamic
calorific values, machine wear and plant modeling errors. model obtained from Step 1) to determine the ultimate
period , the ultimate gains and of the con-
A. Steady-State Triangular Decoupler trolled plant (see the Appendix) corresponding to the
As the time-delay of the turbine and reheater is much power output control loop and the main steam pressure
smaller than that of the boiler , the time-delay in can be control loop, respectively.
ignored and the following approximation in (1) can be obtained: 4) Based on and , determine prescribed ranges
( , ) and ( , ) for the
(2) main steam pressure fuzzy autotuning controller.
In this application, based on the experiments, we
where is the static gain of . Fig. 4 shows the block di- choose , ,
agram of the multivariable feedback control structure of CCS, , .
is the matrix containing the PID 5) From the values of , and , use the tuning
controllers, is the input compensator matrix and is the method proposed in [16] to tune the PI parameters of the
output compensator matrix. power output control loop.
Denote , If and are chosen,
respectively, as
V. BASIC CONTROL LEVEL
and (3)
The boiler-turbine process includes fairly strong nonlinear-
ities due to the difference between the stored energy at each
and substituting (2) into (1), the transfer function matrix of CCS
plant load. To cope with the load-dependent nonlinear dynamic
for the controlled plant becomes
characteristics, in the basic control level the local PID type con-
trollers and decouplers are designed at several load levels based
(4)
on linear control theory firstly according to methods described
in Section III-B, and then gain-scheduling technique according
The strong interaction between the two control loops is reduced,
to the actual load described in Section IV-B is used to obtain the
and the process is now approximately a triangular decoupled
global controllers/decouplers parameters.
system. The parameter of the decoupling compensator matrix
There are three PID-type controllers in the basic control level
can be adapted by the CCS autotuner.
as shown in Fig. 3, each has different functions.
B. Autotuning CCS Parameters 1) PI controller Form and obtained from Step 3) of
the autotuning algorithm, the parameters and can
The autotuner for CCS parameters monitors the control per- be obtained by Ziegler–Nichols method as follows [16]:
formance of the CCS online. If the absolute value of the error
between the set-point and the current value of the main
steam pressure exceed a prescribed threshold for a predeter- (5)
mined period of time, the autotuner automatically retunes the
CCS parameters. Otherwise, the CCS parameters remain un- 2) PD controller A PD controller as shown in Fig. 3 pro-
changed. The autotuning algorithm is given as follows. vides a feedforward compensation to the given load
LI et al.: A NEW COORDINATED CONTROL STRATEGY FOR BOILER-TURBINE SYSTEM http://www.paper.edu.cn
947

under the Voltage-constant running mode. From Figs. 2 is to use autotuning PID controller to implement real-time
or 3, the equivalent transfer function of the system with control.
an integral filter in the generator is given by
A. Fuzzy Autotuning PID Controller
(6) In order to enhance the robustness and control performance of
the main steam pressure loop, a fuzzy autotuning PID controller
With a feedforward controller to compensate the action of is adopted for better performance [9]. By using fuzzy rules based
to , we have on expert knowledge to adjust PID parameters which are ini-
tially determined by classical tuning rules, high-quality control
(7) performance can be expected than that of the PID controllers
with fixed parameters.
where is the controller in the first control loop, and 1) Fuzzy Rules for Tuning PID Parameters: Based on the
feedforward controller can be obtained as step response analysis, human expertise for the process and ex-
(8) tensive simulation studies, a set of autotuning rules for the PID
parameters of the following form are proposed:
This feedforward controller may not be realizable. In if is and is
practice, a PD controller is usually adopted to satisfy this
then is is is (12)
need with the following form:
where and are the current error and its first differ-
(9) ence of the main steam pressure, , , , and represent
a member of fuzzy sets for , , , and , respec-
where the parameters and can be calculated by tively, .
the Ziegler–Nichols method in [16], as follows: For the given prescribed minimum/maximum
of and of in Step 4) of the
(10) autotuning algorithm, the tuning coefficients , and of
, and , are given, respectively, by
and the filter time constant can be determined from
experiments, the engineering rule of thumb value is two (13)
to three times the process time constant.
3) PID controller The PID controller used in this paper is (14)
given as follows:

(11) (15)

According to the Ziegler–Nichols PID tuning rule, when 4,

where the time constant of the first-order lag filter
the integral action of PID controller is moderate. Therefore, by
, and , .
chosen smaller or bigger than 4, a stronger or weaker integral
Through onsite observation and theoretical analysis [5],
action can be obtained.
the main contributing factors for uncertain system perfor-
The fuzzy sets and may be either big (B) or small (S)
mance of the main steam pressure loop are as follows.
and are characterized by the membership functions of natural
• Disturbance caused by frequent unmeasured pulverized
logarithm. The grade of the membership functions and the
coal disturbance under the uncertain coal mill working
variable ( or ) has the following relation:
conditions.
• Parameter uncertainty and variation caused by the uncer- or for big (16)
tainties due to the changing fuel calorific value, machine
wear and the contamination of the boiler heating surfaces. or for small (17)
As the main steam pressure control loop plays a crucial
role in rejecting various disturbances and tracking the Based on the processes characteristics, four singleton member-
power output, high-quality control performance is very ship functions denoting the linguistic variable small (S), middle
important since the change of main steam pressure repre- small (MS), middle (M) and big (B) for the fuzzy sets are de-
sents the energy balance between boiler steam production signed. They are defined as 2.8, 3.4, 4 and
and grid load demand or stored energy in the evaporator. 5.
If the main steam pressure is too high, the stress on the For the system performance to be robust and to achieve fast
plant increases and the plant’s life span will decrease. If it response with small overshoot, the PID controller must be tuned
is too low, the efficiency will decrease. If there exist per- such that the three control actions coordinate with each other as
sistent large main steam pressure variations, the safe op- the process is affected by multiple time-delays and large settling
eration of the unit cannot be guaranteed. When the boiler times. For example, to obtain a desired step response, a big con-
is operated at constant pressure, the main steam pressure trol signal is needed at the beginning to achieve a fast rise time,
dynamics changes with unit load, a better control strategy which requires a big proportional gain and a small derivative
中国科技论文在线
948 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 13, NO. 6, NOVEMBER 2005

TABLE I TABLE II
FUZZY AUTO TUNING RULES OF K FUZZY AUTO TUNING RULES OF K

TABLE III
gain. Also, since processes have large time-delay, a small inte- FUZZY AUTO TUNING RULES OF 
gral gain is desired to reduce the overshoot. When the output
response is near the set-point, the proportional gain and integral
gain should be changed from large to small and from small to
large, respectively, to make the controlled output converge to
the set-point quickly. The autotuning rules for , and
are given in Tables I–III, respectively.
2) Fuzzification Strategy and Fuzzy Inference: In fuzzifica-
tion, the shape of membership function characterizes the intu-
ition of converting the crisp value into linguistic value to fit the
human thinking process. Gaussian-shaped membership function
is chosen as the membership function of the antecedent part
or in (12) as it better fits human intuition. Similar functions given in (8) and (9). Then the defuzzification yields
to the triangle partition system with evenly space (TPE) [18], the following results:
Gaussian-shaped membership function have the following fea-
tures: 1) the membership function is symmetrical about its cen-
tral value, 2) all membership functions have the same shape, and (20)
3) the space between the central values of two adjacent members
are equal and these constitute the GPE used in our system.
The Gaussian membership function is defined as
(21)
(18)
others
where is the center value of the membership function of
. Point is a unique element that has membership value (22)
1 in , this guarantees four rules without zero contribution at
any one time.
According to the fuzzy autotuning rule given in (12), where or is the value of or corresponding to the
and Tables I–III, the crisp sampling value or grade for the th rule . Once , and are obtained,
is first fuzzified into linguistic value based on the previous the PID controller parameters , and can be autotuned
fuzzification strategy by seven reference membership functions online based on (13)–(15).
denoting the linguistic variables: negative big (NB), negative
middle (NM), negative small (NS), zero (Z), positive small B. Gain-Scheduling of the PID Type Controllers and
(PS), positive middle (PM) and positive big (PB), respectively. Decouplers
Then, the crisp value of the th rule given in (12) is If it is known how the dynamics of a process change with the
obtained by the product of the membership function values of operating conditions of the process, it is possible to change the
and controller parameters accordingly, known as gain-scheduling.
(19) A measurable process variable, which is descriptive of the oper-
ating condition and used to adjust the controller parameters, is
where is the membership function value of the refer- known as a scheduling variable. For the boiler-turbine process,
ence fuzzy set given a value of , and is the the actual load is chosen as the scheduling variable.
membership function value of the reference fuzzy set given A set , containing m values of
a value of . Based on , the values of and for the scheduling variable is chosen and arranged according to:
each rule are determined from their corresponding membership for ). For each value of
LI et al.: A NEW COORDINATED CONTROL STRATEGY FOR BOILER-TURBINE SYSTEM http://www.paper.edu.cn
949

in the set a linear model ( TABLE IV

, 2) is identified by using identification methods described in MODEL PARAMETERS OF 210 MW
Section III-B and for each model parameters of the local PID
type controllers and decouplers are tuned according to methods
described in Section III-B.
By assuming that the parameters of the PID type controllers
and decouplers change linearly between the two load levels,
the parameters can be gain-scheduled between the frozen oper-
ating points by linear interpolation

(23)
according to the actual load , where . TABLE V
MODEL PARAMETERS OF 270 MW

VI. INDUSTRIAL IMPLEMENTATION AND APPLICATION RESULTS

The proposed coordinated control strategy described in this
paper is realized by using the software development system of
the FOXBORO I/A Series DCS and has been developed and
successfully applied in the control of a 300 MW boiler-turbine
unit, i.e., Unit 1 of Yuanbaoshan Power Plant in China.

A. Identifying the Plant Models and Autotuning Controller

Parameters
To design an efficient control system for such a complicated
plant reliably, open-loop identification of the control plant was
performed under different coal quality and load levels by using
identification methods described in Section III-B. For example,
parameters at two typical operating conditions of 210 MW
(70% of full load) and 270 MW (90% of full load) and with
calorific value of the coal in the range of 3300 to 3600 kcal/kg
are obtained and listed in Tables IV and V. The form of these
linear models is given in (1) where
( 1, 2).
Remark 1: The identified plant models have been exten- operating conditions described previously were designed for in-
sively validated, including field dynamic experiments carried stance as follows.
out before the control system have been designed, the open-loop • The steady triangular decoupler parameter was tuned
identification by using the methods described in Section III-B from the steady-state plant model parameters as 0.936 and
and through comparisons between simulation results and actual 0.962 corresponding to the 210 MW and 270 MW oper-
plant data. From the validations, it has been concluded that ating conditions, respectively.
the plant models reflect the process dynamics in a satisfactory • 1 of (6) was selected.
manner, so that the initial design of the control system can be • The desired critical period and critical gains ( , and
performed on the plant models. ) were determined as (727 s, 0.269 and 2.304) and
Remark 2: The identified plant models are only accurate at (556 s, 0.192 and 2.046) corresponding to the 210 MW
the typical operating conditions. In order to avoid the disadvan- and 270 MW operating conditions, respectively.
tages of self-tuning control based on rational transfer function • The PI controller parameters ( , ) of the power output
parameter estimation directly, increase the robustness and facil- control loop were tuned as (1.019, 60.58 s) and (0.598,
itate commissioning, self-tuning control method from the fre- 49.86 s), respectively, for the two operating conditions ac-
quency domain approach [19] is adopted by the estimation of the cording to the values of , and based on spec-
ultimate point of the controlled plant. Simulation studies show ified phase margin and specified amplitude
that the changes of the ultimate point parameters of the con- margin 3.
trolled plant are small when there are large changes in the pa- • Based on simulation studies and real-time experiments,
rameters of the plant transfer function model. The results point the ranges for ( , ) and ( , )
to the robustness of the proposed self-tuning control method. are obtained as , ,
To illustrate the design procedure of the proposed CCS, the , ,
local decoupler/controllers initial parameters of the two typical respectively.
中国科技论文在线
950 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 13, NO. 6, NOVEMBER 2005

Fig. 6. CCS at load change rate of 3 MW/min.

Fig. 7. CCS at load change rate of 10 MW/min.

After the local PID type controllers/decouplers are designed

at several typical load levels based on linear control models, the
controller/decouplers parameters of other operating conditions
are gain-scheduled according to method given in Section IV-B.

B. Real-Time Application and Discussion

After commissioning, the proposed CCS has been put into
service in the 300 MW commercial power-generating unit for
more than three years. Here, several real-time operation results
are recorded and used as examples to demonstrate the effective-
ness of control strategy.
Case 1: An essential experiment for the boiler-turbine unit
is increasing the load at the rate of 3 MW/min to test the control Fig. 8. CCS at load change rate of 9 MW/min.
performance, before the CCS can be put in operation. Fig. 6
shows the operation results for a 180-min period, where the Case 3: Fig. 8 shows the operation result under coordinated
main steam pressure and the power output increase control when the unit is decreasing the load over a wide range
from 13.56 MPa to 18.06 MPa and from 185 MW to 285 MW, at the rate of 9 MW/min from 287 MW to 197 MW. As can be
respectively. The controlled variables showed fast response with seen, the principal controlled variables of the unit are all kept
small tracking error and good steady-state performance for both near their setpoints and good load following is achieved, so the
the main steam pressure and the power output, while the main efficiency of the unit is enhanced.
steam temperature still maintains the set-point of 545 C. Case 4: The 4-h operation results of the proposed CCS
Case 2: Load changes over a wide range between 185 MW and manual operation under high load condition are compared
and 285 MW at the allowable maximum rate of 10 MW/min. In during commissioning to validate the proposed control scheme.
this case, the main steam temperature will be severely affected As shown in Figs. 9 and 10, the main steam pressure and the
under conventional control strategy. Fig. 7 shows three typical power output under the proposed control strategy are running
operating conditions within the specified operation range; much smoother than those of manual operation under the full
1) decreasing the set-point of two control loops and load 300 MW. In manual control mode, the frequent strong
during 20–40 min; 2) setting and as pulse signals during unmeasured pulverized coal disturbance results in the main
90–100 min; and 3) increasing and during 150–180 min. steam pressure exceeding the prescribed upper limit for safe
The controlled variables and , as well as main steam operation and cause the high-pressure bypass valve to open
temperature are all kept within acceptable limits near resulting in wasted energy. Under automatic control mode, the
their set-points and good tracking performance is obtained. system has a very good disturbance rejection property, safe and
Since the maximum rate of load changes in the power plant is economic operation of the unit is guaranteed.
6 MW/min and the system has good tracking performance even Remark 3: This paper is based on the practical application
at is 10 MW/min (150–160 min.), the proposed control strategy project which had been carried out at Yuanbaoshan Power
satisfies the load following requirement. Plant, China. While many of the control loops were regulated
LI et al.: A NEW COORDINATED CONTROL STRATEGY FOR BOILER-TURBINE SYSTEM http://www.paper.edu.cn
951

Fig. 11. Closed-loop TITO control system.

2) both the main steam pressure and the electric power

output have fast responses with low overshoot and good
Fig. 9. CCS under manual control. steady-state performance;
3) the main variables of the unit including the main steam
temperature are all kept within acceptable limits near
their set-points and good tracking performance even
under load-variation over a wide range.
The CCS proposed can be easily implemented in other coal-
fired boiler-turbine unit of the power plants without much mod-
ification. The research work on the extension of the technology
and an adaptive predictive control for the CCS of power plant is
currently under investigation.

APPENDIX
RECURSIVE LEAST SQUARES IDENTIFICATION ALGORITHM
The simple identification approach is on the basis of a MIMO
process under decentralized control. To simplify our derivation,
we adopted two inputs and two outputs control system (shown
Fig. 10. CCS under automatic control. in Fig. 11).
Where , and controllers, noises and process transfer
in automatic mode, the main steam pressure loop and the power functions, respectively, the notation , , 1, 2, are
output loop have had to be controlled by operators because of used in both and domain. In general, and could
the limitations of the conventional CCS. As it is difficult to be any type of controllers that make the closed-loop system
guarantee safe operation of the plant using a conventional CCS, stable. Without loss of generality, assumed that the controllers
such studies were not conducted, and, hence, no comparative and are proportional type. The fundamental relation-
results are available between the proposed approach and the ship between error signals and transfer function outputs for the
conventional CSS. Hence, the results of the proposed approach system are expressed as
and manual operation of the plant is presented, to demonstrate
the improvement in performance with the proposed CSS. and

VII. CONCLUSION Assumed that the process initially test at a steady state with
initial set point, error, and output variables. Notes that , , ,
A new coordinated control strategy for the boiler-turbine unit , and , respectively, and then , .
in a power plant was proposed in this paper to improve the To identify the process parameters, the test involves the fol-
system performance under the load following mode. The system lowing two steps.
consists of two levels with a fuzzy inference system for auto- 1) When kept fixed, make a step change from to ,
tuning the PID controller. The CCS and fuzzy autotuning PID record the error signals for the two loops, until the new
controllers have been implemented in a 300 MW boiler-turbine
steady state is reached at . is defined
unit in China for more than two years. The system has been per-
as the maximum settling time of all loops (See Fig. 2).
forming very well after three years of fine-tuning. In summary,
The incremental equation from the second steady state to
we have adverted the following:
the new state becomes
1) the CCS had successfully replaced the manual operation
in loop coordinate control and the performance has been
(A1a)
very robust;
中国科技论文在线
952 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 13, NO. 6, NOVEMBER 2005

decentralized controllers are proportional controllers, that is

(A1b) both and are constants, the relation between the original
system input–output and the decentralized identification system
2) Make a step change in from to , while keeping are given as.
as before, record the error signals for the two Consider a first-order plus time delay and a second-order plus
loops, until the new steady state is arrived at time delay system, the solutions for 1 and 2 are given
. Again, the incremental equation from the third as follows.
steady state to the new state can be written as 1) 1, each loop output results becomes
(A2a)

(A2b) (A5)
which can be written into the compact form
Combine (A1a), (A1b), (A2a) and (A2b) into matrix form
(A3a)
where where

2) 2, (A5) becomes

and

(A3b)

(A6)
Because of ,
thus, the matrix is nonsingular if
. From (A1) and (A2), it can be proved that It is again expressed as

where

Therefore, for a bound-input–bound-output (BIBO)

stable system, , in Laplace domain.
Now that is nonsingular, can be solved by
(A4)
Substituting (A1b), (A2b) and (A3b) into (A4), we obtain the Equations (A5) and (A6) can be solved by the Least
equation shown at the bottom of the page. Squares methods for each transfer functions 1, 2
Consequently, the problem of identification of coupled and 1, 2, to form the regression form
closed-loop MIMO system is transformed into the identi-
fication of four single open-loop problems. Suppose that
LI et al.: A NEW COORDINATED CONTROL STRATEGY FOR BOILER-TURBINE SYSTEM http://www.paper.edu.cn
953

ACKNOWLEDGMENT
The authors would like to thank the anonymous reviewers
for their helpful comments and constructive suggestions with
regard to this paper.

REFERENCES
[1] C. Maffezzoni, “Boiler-turbine dynamics in power-plant control,” Con-
trol Eng. Pract., vol. 5, no. 3, pp. 301–312, 1997.
[2] S. Goto, M. Nakamura, and S. Matsumura, “Automatic realization of
human experience for controlling variable-pressure boilers,” Control
Eng. Pract., vol. 10, no. 1, pp. 15–22, 2002.
[3] T. Chai, H. Yue, and Y. Bai, “Intelligent control of coal-pulverizing sys-
tems with ball-tube mill,” in Proc. IFAC 14th Triennial World Congr.,
Beijing, China, 1999, pp. 97–102.
[4] P. Zgorzelski, H. Unbehauen, and A. Niederlinski, “A new simple de-
centralized adaptive multivariable regulators and its application to mul-
tivariable plants,” in Proc. IFAC 11th World Congr., vol. 4, Tallinn, 1990,
pp. 226–231.
[5] H. Liu, “A Study on advanced control techniques application in the large
thermal power generating unit,” Doctoral dissertation, AU: WHAT DE-
PARTMENT? Northeastern Univ., China, 2000.
[6] J. A. Rossiter, B. Kouvaritakis, and R. M. Dunnett, “Application of gen-
eralized predictive control to a boiler-turbine unite for electricity gener-
ation,” Inst. Elect. Eng. Proc.—D, vol. 138, no. 1, pp. 59–67, 1991.
where , [7] B. W. Hogg and N. M. El-Rabie, “Generalized predictive control of
. steam pressure in a drum boiler,” IEEE Trans. Energy Convers., vol. 5,
no. 3, pp. 485–492, Sep. 1990.
Its Least Squares estimation for are [8] , “Multivariable generalized predictive control of a boiler system,”
IEEE Trans. Energy Convers., vol. 6, no. 2, pp. 282–288, Jun. 1991.
[9] W. Wang, H. X. Li, and J. T. Zhang, “Intelligence-based hybrid control
(A7) for power plant boiler,” IEEE Trans. Contr. Syst. Technol., vol. 10, no.
2, pp. 280–287, Mar. 2002.
[10] G. Zhang, C. Shao, and T. Chai, “A new method for independently
Once are found from (A7), , and can be recov- tuning PID parameters,” in Proc. 35th IEEE Conf. Decision Control,
ered from Kobe, Japan, 1996, pp. 2527–2532.
[11] Q. Bi, W. J. Cai, and Q. G. Wang et al., “Robust identification of first-
order plus dead-time model from step response,” Control Eng. Pract.,
vol. 7, no. 1, pp. 71–77, 1999.
[12] , “Advanced controller auto-tuning and its application in HVAC
systems,” Control Eng. Pract., vol. 8, no. 5, pp. 633–644, 2000.
[13] E. Poulin, A. Pomerleau, and A. Desbiens et al., “Development and eval-
for , and , , , and from uation of an auto-tuning and adaptive PID controller,” Automatica, vol.
32, no. 1, pp. 71–82, 1996.
[14] Q. G. Wang, C. C. Hang, and Q. Bi, “Process frequency response es-
timation from relay feedback,” Control Eng. Pract., vol. 5, no. 9, pp.
1293–1302, 1997.
[15] R. Bandyopadhyay and D. Patranabis, “A fuzzy logic based PI auto-
tuner,” ISA Trans., vol. 37, pp. 227–235, 1998.
[16] K. K. Tang, Q. G. Wang, and C. C. Hang, Advances in PID Con-
for 2, respectively. trol. New Yprk: Springer-Verlag, 1999.
[17] X. Liu, T. Chai, and H. Liu, “Fuzzy logic strategy for boiler control,”
Acta Automatica Sinica, vol. 24, no. 4, pp. 534–538, 1998.
[18] T. Sudkamp and R. J. Hammel, “Interpolation, completion, and learning
APPENDIX fuzzy rules,” IEEE Trans. Syst., Man, Cybern., vol. 24, no. 2, pp.
ZIEGLER–NICHOLS FREQUENCY RESPONSE METHOD [16] 332–342, Feb. 1994.
[19] W. K. Ho, C. C. Hang, W. Wojsznis, and Q. H. Tao, “Frequency domain
The Ziegler–Nichols frequency method is based on using the approach to self-tuning PID control,” Control Eng. Pract., vol. 4, no. 6,
pp. 807–813, 1996.
controller connected as a proportional controller, the experiment
is carried out in the following procedures.
1) Connect a controller to the process, set the parameters
so that control action is proportional, i.e., , and
0. Shaoyuan Li (SM’05) was born in 1965. He received
2) Increase the control gain slowly until the process starts to the B.S. and M.S. degrees from Hebei University of
Technology, Tangshan, China, in 1987 and 1992, re-
oscillate. spectively, and the Ph.D. degree from the Department
3) The gain when this occurs is , and the period of the of Computer and System Science, Nankai University,
oscillation is . Tianjin, Beijing, China in 1997.
He is currently a Professor at the Institute of Au-
4) Determine the PID controller parameters according to the tomation, Shanghai Jiao Tong University, Shanghai,
Ziegler–Nichols method as shown in the figure at the top China. His research interests include fuzzy systems
of the page. and nonlinear system control.
中国科技论文在线
954 IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 13, NO. 6, NOVEMBER 2005

Hongbo Liu was born in 1964. He received Yeng-Chai Soh received the B.Eng. degree in elec-
the Ph.D. degree from the Research Center of trical and electronic engineering from the University
Automation, Northeastern University, China, in of Canterbury, New Zealand, in 1983 and the Ph.D.
2000. He is currently working toward the Ph.D. degree in electronic engineering from the University
degree at the Institute of Automation, Shanghai Jiao of Newcastle, Australia, in 1987.
Tong University, Shanghai, China. He is currently a Professor in the School of
His research interests include complex thermal Electrical and Electronic Engineering, Nanyang
process modeling and control, fuzzy control, and Technological University, Singapore. Since 1995, he
adaptive control. has been the Head of the Control and Instrumentation
Division. His current research interests are in the
areas of robust system theory and applications.

Wen-Jian Cai was born in 1957. He received the Li-Hua Xie received the B.E. and M.E. degree in
B.S. and M.S. degrees from Harbin Institute of electrical engineering from the Nanjing University
Technology, Harbin, China, in 1980 and 1983, of Science and Technology, Nanjing, China, in
respectively, and the Ph.D. degree in systems engi- 1983 and 1986, respectively, and the Ph.D. degree
neering, Oakland University, Rochester, MI, in 1992. in electronic engineering from the University of
He is currently an Associate Professor in the Newcastle, Australia, in 1992.
School of Electrical and Electronic Engineering, He is currently an Associate Professor in the
Nanyang Technological University, Singapore. His School of Electrical and Electronic Engineering,
research interest includes advanced process control, Nanyang Technological University, Singapore.
fuzzy logic control, robust control, and estimation His current research interests include optimal and
techniques. robust control.