CN102753789B - The method and apparatus that steam in steam regulation power equipment produces - Google Patents

Abstract

The present invention is based on a kind of for regulating the method being produced steam (16) by the water supply (10) in the vaporizer of apparatus of steam power plants (6), wherein state regulator (30) calculates the multiple medium state in vaporizer (6) by visualizer (42) and therefrom determines quality of water supply stream (m _s) as regulating parameter.In order to realize, to the stable of vapor (steam) temperature and accurate adjustment, advising that described state regulator (30) is linear quadratic regulator.

Description

The method and apparatus that steam in steam regulation power equipment produces

Technical field

The present invention relates to a kind of for regulating by the method for the confession water generates steam in the vaporizer of apparatus of steam power plants, wherein in the first regulating system, state regulator calculates the multiple medium state in vaporizer by visualizer and therefrom determines the regulating parameter of quality of water supply stream as the first regulating system.

Background technique

The efficiency of apparatus of steam power plants is along with the temperature of the steam produced in power station boiler and improve along with the stability of the quality of the steam provided after evaporator unit.Steam in apparatus of steam power plants produces general from being preheated high pressure pre-heater (also referred to as vapor economizer) and then being undertaken by the water supply of evaporating in vaporizer.Water supply was set to high pressure by raw water service pump at this and is driven through high pressure pre-heater and vaporizer before high pressure pre-heater.

The adjustment of evaporating temperature is after the evaporator by being set to regulating parameter carry out by the mass flow of the water supply be transported in vaporizer.The adjustment behavior with the vapor (steam) temperature of this regulating parameter is lentissimo, thus the adjustment of quality of water supply stream just acts on temperature to be regulated after a few minutes.In addition, temperature to be regulated is subject to the strong impact of countless interference, and these interference are such as the replacings etc. of blowing ash, fuel in load variations, boiler.Accurate temperature regulates because these reasons are difficult to realize.

Summary of the invention

Task of the present invention is that a kind of method is described, utilizes the method can not only accurately but also stably steam regulation temperature.

This task solves in the following manner, and namely state regulator is linear quadratic regulator according to the present invention.Such linear quadratic regulator (LQR) can comprise the optimum state feedback of linear quadratic.At this, can determine the parameter of this regulator in such a way, the quality standard namely can exchanging joint quality is optimized.Not only accurate but also stable adjustment can be realized thus.

The present invention at this based on following consideration: multiple (part is immeasurablel) state of feeding back when status adjustment is for determining regulating parameter (or regulator regulated signal).For current applicable cases, this means to be used in the state along on multiple positions of vaporizer in the algorithm, as temperature, pressure, enthalpy or other status parameter.But due to these state immeasurabilities, therefore need so-called observer circuit, can estimate or calculate the state of the needs that can be characterized by status parameter by this observer circuit.Concept " estimation ", " calculating " and " determination " synonymously use below.The advantage of this concept is, can very fast and accurately react the interference acting on vaporizer.

Apparatus of steam power plants is the equipment utilizing steam power to run.It can be or comprise steam turbine, steam course equipment or any equipment that other runs with the energy from steam.As vaporizer, can be understood as any system of being evaporated by water below, wherein preheater, especially high pressure pre-heater are includable.Medium can be supply water, steam or by supplying water and the mixture of vapour composition.Medium state (being also called state simply below) can be other state of energy, temperature, pressure, enthalpy or this medium.

As state regulator, can be understood as regulating loop regulating parameter regulated based on estimated state (such as with the form of state space display) below.At this, estimate regulate the one or more state in section and be transported to again by visualizer, namely feed back to adjustment section-or regulator.The feedback forming regulating loop together with adjustment section can be undertaken by visualizer, and visualizer can replace measuring device thus.The state of visualizer observation or estimating system is the state of the medium in vaporizer in this case, and can comprise state difference equation, output equation and visualizer vector.The output of visualizer can compared with regulating the output of section.Difference can act on state difference equation by visualizer vector.In addition advantageously, visualizer and state regulator independently work.

It is suitable that, state regulator uses the state leaving the steam of vaporizer as regulating parameter, as the enthalpy of vapor (steam) temperature or steam.Quality of water supply stream is advantageously used as regulating parameter.

Launch in favourable enforcement of the present invention, the rating value of quality of water supply stream is transmitted to the regulator of the second regulating system for adjustment quality of water supply stream.Described rating value can be used as regulating parameter by this regulator.As the regulating parameter of the second regulating system, the position of the rotating speed of raw water service pump, valve (such as in water supply conduit) can be used directly or indirectly or be applicable to regulate other parameter of quality of water supply stream.

In addition advantageously, use the enthalpy of medium as status parameter in order to calculation medium state.It is suitable that, use multiple state and use its multiple enthalpy thus.Steam parameter (as enthalpy and/or pressure and temperature) should remain the value of expectation according to loading condition, and should be adjusted in load variations phase.The advantage of enthalpy status adjustment, namely, the product of enthalpy or enthalpy and other parameter (as quality of water supply stream) is used to be that status adjustment reaches higher regulation quality and adjustment is faster as the advantage of state.Method and technology also obtains following advantage: this process is so regulated suitably, make at the faint overheated steam of vaporizer end effusion, this steam is close to the saturated vapour limit.Utilize the pressure (such as in sliding pressure runs) of change, the evaporation terminal of formation wet vapor or saturated vapour point may be caused when observed temperature also to change.When enthalpy is used as status parameter, pressure does not need to observe together clearly because enthalpy by temperature and force combination in a parameter.

Advantageously, the deviation of absolute enthalpy and enthalpy rating value is used as status parameter.Can 0 be adjusted in the state of the equilibrium and simplify mathematical problem thus.

LQR method relates to linearly adjust problem.By measured temperature and temperature rating are scaled enthalpy, mathematics can be regulated problem linearization and reach thus when using enthalpy state and more simply calculate, because there is linear correlation between input enthalpy and effusion enthalpy.It is suitable that, this conversion is carried out when using such as measured vapor pressure by corresponding water/steam chart relation.

But produce following problem when regulating vaporizer section by status adjustment, although the state namely at evaporator inlet place can be described by enthalpy, but the enthalpy at evaporator inlet place can not be arranged, because the pressure and temperature supplied water can only be time strategic point change and be not suitable as regulating parameter.Therefore it is suitable that quality of water supply stream be used as regulating parameter and be multiplied with state when computing mode.

But, act on the enthalpy at regulating parameter-evaporator inlet and evaporator outlet place in a non-linear fashion for current, although thus employ enthalpy, adjustment problem is also nonlinear.In order to solve this problem, it is suitable that use linearization when computing mode.Advantageously suppose in current situations, state is the mobile deviation range around an operation point only.System can be assumed to linear in this deviation range be determined in advance suitably.

This linearization for an only operation point state and be significant for the deviation range around this operation point.If the state of reality shifts out this deviation range, then linearization will cause disadvantageous result.Therefore advantageously operation point is upgraded.This it is suitable that carries out in the following manner, namely by adopting measured value to upgrade described operation point.It is suitable that, measured value is the current measurement value that the medium parameter by measuring current existence gathers as pressure, temperature etc.State computation based on operation point can mate with current media state.Nonlinear regulating system can being used, coming this regulating system linearization by adopting current measurement value.Realized the adjusting function of very robust by linearization, that is quality of regulation no longer depends on the current point of operation of equipment.

Another Advantageous embodiments regulation of the present invention, the regulating system of state regulator comprises the matrix equation of such as feedback matrix form, and dielectric value measured during producing steam is used to calculate described matrix equation.Thus such as can carry out feedback of status by the matrix equation using current measurement value to determine at least in part by its parameter.By use current measurement value, such as, at feedback matrix in line computation, regulator can mate with the operating conditions of reality all the time.Automatically the change depending on load of dynamic evaporation device performance can be considered thus.The raising of the robustness regulating algorithm can also be realized by this step.Based on the fact regulating the unusual robust of algorithm, only need to arrange considerably less parameter when starting.Therefore, starting time and Start-up costs remarkable minimizing compared with all methods known at present.

Advantageously, compute matrix equation is carried out by the control technique of apparatus of steam power plants.At this, this control technique can be the control system controlling this apparatus of steam power plants in the day-to-day operation of apparatus of steam power plants.In order to keep the mathematics assembly of this control technique simple, advantageously matrix equation is converted to the difference equation of one group of scalar.Can be realized by reverse integral in time the relatively simple integration of matrix equation.Due to not available from the information in future in the ideal case, if therefore with the DIFFERENCE EQUATIONS integration (this stably cause same fixing solution) of contrary symbol to this scalar, the integration with reverse integral equivalence can be realized.

In Advantageous embodiments of the present invention, visualizer is the Kalman filter of linear second state Feedback Design.Linear quadratic regulator is called LQG(linear quadratic gaussian with cooperating of Kalman filter) regulator or LQG algorithm.

Advantageously, visualizer calculates the heat entering medium in vaporizer.This heat can be defined as interference parameter and use in adjustment algorithm.At this not only along the enthalpy of vaporizer or the parameter therefrom derived, and interference parameter can be defined as state in addition, and especially can estimate by visualizer or determine.The interference directly acting on vaporizer represents in the following manner, and the preheating interval namely in vaporizer changes.By the such observation to interference parameter, but can to carry out very fast corresponding interference, the reaction of robust simultaneously accurately.

In addition the present invention relates to a kind of equipment for regulating from the confession water generates steam the vaporizer of apparatus of steam power plants, there is regulating system, this regulating system comprises visualizer and state regulator, and described state regulator prepares to be used for calculating the multiple medium states in vaporizer by visualizer and therefrom determining the regulating parameter of quality of water supply stream as the first regulating system.

Proposed states regulator is linear quadratic regulator.Accurate and stable adjustment can be realized.

Advantageously, described equipment be designed to perform in above-mentioned advised method step one, multiple or whole.

Accompanying drawing explanation

The present invention is elaborated by embodiment illustrated in the accompanying drawings.

Fig. 1 illustrates a fragment of the steam power station with vaporizer,

Fig. 2 illustrates the plan regulating cascade,

Fig. 3 illustrates the model of vaporizer,

Fig. 4 illustrates the linearity range model on the basis as design of Regulator,

Fig. 5 illustrates the structure of visualizer, and

Fig. 6 illustrates the general view of controller structure.

Embodiment

Fig. 1 illustrates the schematic diagram of a fragment of the thermoelectricity plant with apparatus of steam power plants, and this apparatus of steam power plants comprises steam turbine 2, boiler 4, vaporizer 6 and superheater 8.Boiler 4 is by heat output to vaporizer 6, and water supply 10 is through-flow in vaporizer 6, and described water supply 10 pumps to vaporizer 6 by raw water service pump 12 and described water supply absorbs heat.Can regulate water supply stream by valve 14.

Water supply 10 is evaporated in vaporizer 6 by absorbing heat, and the steam 16 produced flows to superheater 8 further so that overheated one-tenth live steam is also then transported to steam turbine 2 there.In order to the temperature of steam regulation 16, regulate for current by valve 14 and/or raw water service pump 12, wherein the specified stream of water supply 10 before vaporizer 6 is regulating parameter, and valve location and/or pump power are regulating parameter.

The temperature T of water supply 10 measured by temperature transducer 18 and pressure transducer 19 _wor pressure p _w, and the actual confession current m before vaporizer 6 measured by sensor 20 _i.The vapor (steam) temperature T of the steam 16 after vaporizer 6 measured by temperature transducer 22 and pressure transducer 24 _dor vapor pressure p _d.

Vaporizer 6 can comprise unshowned preheater.But this is unimportant to the present invention, and the concept of " vaporizer " it should also be understood that the system for being made up of the vaporizer with preheater below.

Vaporizer 6 is pump circulation steam generators, is wherein forced the circulation of current or vapor stream by feed water pump 12.Supply water and 10 can successively flow through for water preheater and evaporation component at this, especially also by superheater 8, thus supply water and 10 be heated to saturated-steam temperature, evaporation and overheatedly carry out in the circulating cycle continuously.Rotating cylinder is not needed at this.Especially vaporizer 6 is parts of Benson boiler.It can enter critical range, wherein supplied water 10 to be set to pressure more than 230 bar by raw water service pump 12.Can regulate quality of water supply stream according to load.

Fig. 2 illustrates to have first or external regulation system 26 and second or the adjustment cascade of internal regulation system 28.External regulation system 26 comprises linear quadratic regulator 30, especially LQG regulator.As input parameter, to the actual confession current m measured by its conveying _i, supply water 10 measured temperature T _w, steam 16 measured temperature T _dwith measured pressure p _dand the rated temperature T of steam 16 after vaporizer 6 _s.The rated temperature T of steam 16 _sit is the regulating parameter of regulator 30.The nominal-mass stream m of water supply 10 _sexported by regulator 30 as regulating parameter.

This nominal-mass stream m _sthe scheduled rating value of regulating loop 32 as regulating parameter giving internal regulation system 28.Measured supplies current m _iit is the regulating parameter of regulating loop 32.The power of position and/or raw water service pump 12 that regulating loop 32 has regulating valve 14 is as regulating parameter.

Regulator 30 is not directly act on this process by regulating mechanism, but by the rating value m of quality of water supply stream _sbe transmitted to the regulating loop 32 of bottom, it forms the cascade be made up of external regulation system 26 and internal regulation system 28 together with this regulating loop thus.The temperature T measured by water supply 10 before vaporizer 6 _wand pressure p _wthe additional information that regulator 30 needs, for the specific enthalpy h of the water supply 10 determined before vaporizer 6 ₁.Enthalpy h ₁can be determined by water-steam chart.By vapor pressure p _dwith vapor (steam) temperature T _dcalculate the specific enthalpy h of the steam 16 after vaporizer 6 ₂.

Fig. 3 illustrates the model of the evaporator section in vaporizer 6, and this evaporator section is divided into three first-order lag unit 34, thus in their series circuit, obtain the lag characteristic on three rank.Three delay cells can be the PT realized by reverse feedback integrator 36 respectively ₁unit.The time constant of these delay cells depends on load and becomes larger along with the decline of load, and vice versa.To the x that does well after each delay cell 34 _i, wherein i=1,2,3, wherein state x ₁illustrate and export enthalpy h ₂.Input state is characterized by the input enthalpy h1 of evaporator section.Two intermediateness x ₂, x ₃be arithmetic type and be the immeasurablel state estimated by visualizer.All state x _iit is the parameter depending on the time.

There is enthalpy h ₁water supply 10 flow in evaporator section.In principle, this enthalpy h ₁the regulating parameter of first or external regulation system 26 can be used as, because prove the hypothesis of the linear performance of evaporator section by enthalpy instead of temperature.But, enthalpy h ₁almost can not arrange, because the pressure p supplied water _wwith temperature T _walmost can not be adjusted in scope and speed and be enough to be used as regulating parameter.

In order to solve this problem, the actual mass stream m of water supply 10 _iwith enthalpy h ₁be multiplied, thus obtain power from this product.This power can regulate by raw water service pump 12 and/or valve 14 simply and can be used as regulating parameter thus.Due to enthalpy h ₁substantially be constant, the actual mass stream m of therefore water supply 10 _iregulating parameter can be used as individually.

Correspondingly, in the dynamic model shown in Fig. 3 in each delay cell 34 respectively by m _ibe multiplied with current enthalpy, as shown in by multiplier 38, thus form power as parameter.By this power three delay-level 34 each in respectively with supposition combustion power Q _f1/3 to be added, thus by whole combustion power Q _fbe incorporated in the dynamic model of whole evaporator section.

This power and being multiplied with time quantum G, this power and denominator comprise the time constant of delay, such as PT under full load conditions ₁the time constant t of the delay of unit.In addition, G=(mt) ^-1denominator comprise quality of water supply stream m, such as, quality of water supply stream under full load conditions, thus after time quantum G per time all there is specific enthalpy.This specific enthalpy is respectively by integrator 36 integration in each delay cell 34, thus enthalpy exists as a result.

This enthalpy is deducted from the input enthalpy of respective delay cell 34.As the state x after three delay cell 34 _iequation obtain:

${\stackrel{\·}{x}}_1=\frac{1}{\mathrm{mt}}(\frac{Q_F}{3}+m_1(x_2-x_1))$

${\stackrel{\·}{x}}_2=\frac{1}{\mathrm{mt}}(\frac{Q_F}{3}+m_2(x_3-x_2))$

${\stackrel{\·}{x}}_3=\frac{1}{\mathrm{mt}}(\frac{Q_F}{3}+m_3(h_1-x_3)).$

State x ₁export enthalpy h ₂.Can find out, when enthalpy difference is by delay cell 34 and for current m _ibeing multiplied, to add 1/3 of combustion power QF be 0, and enthalpy difference is multiplied by for current m _iand Q _fduring/3 maintenance balance, state x is constant, namely its derivative zero.In this case, under system is in starting of oscillation state and be in thus supply water conveying and heating balance in.

These three equations are not linear, because state x _iwith confession current m _ibe multiplied.This is correct, because the item of the instability of ignition heat should non-linearly be mapped.Ignition heat non-linear in state model, by state x in the visualizer described in detail at Fig. 5 or rather _iwith confession current m _ibe multiplied to simulate.Thus, for current m _ichange and unstable combustion power Q _fon the contrary, as compensating its reaction device.Therefore, for current m _ibe used as the regulating parameter of the first regulating system 26.

In order to LQ regulator or LQG regulator can be used, by linearization, this Nonlinear System of Equations must be converted to linear equations.For this reason, state and input be first expressed as fixed value and with the deviation of fixed value and.Stationary state is obtained by Nonlinear System of Equations, and its method makes the time-derivative of state equal 0.This means, the state in system no longer time of origin change and these states are in fixing state of rest.In addition, stationary state is defined as rated condition.

Correspondingly, stationary state following formula is set up:

$h_2=h_1+\frac{Q_F}{m_s},$

Wherein m _sbeing the quality of water supply stream expected, utilizing it to realize this stationary state, therefore supplying current in the stationary state just in time greatly to making for current output enthalpy h after the evaporator ₂heat conveying Q is absorbed when constant _f.By the regulated value m of conversion acquisition first regulating system _s:

$m_s=\frac{Q_F}{h_2-h_1}$

Then further in order to linearization, assuming that state and the input only mobile deviation range around operation point.This system can be assumed to linear in this operation point thus.As work range selection rated condition, u represents the input of system:

x _i=x _i，soll+Δx _i

u=m _s+Δu

At the product of supposition deviation, i.e. Δ u Δ x _ivery little and in insignificant situation, produce following Linearized state equations:

$\mathrm{\Δ}{\stackrel{\·}{x}}_1=\frac{1}{m\·T}(\frac{Q_F}{h_{2s}-h_1}\·({\mathrm{\Δx}}_2-x_1)-\frac{h_{2s}-h_1}{3}\·\mathrm{\Δu})$

$\mathrm{\Δ}{\stackrel{\·}{x}}_2=\frac{1}{m\·T}(\frac{Q_F}{h_{2s}-h_1}\·({\mathrm{\Δx}}_3-x_2)-\frac{h_{2s}-h_1}{3}\·\mathrm{\Δu})$

$\mathrm{\Δ}{\stackrel{\·}{x}}_3=\frac{1}{m\·T}(\frac{Q_F}{h_{2s}-h_1}\·\left({\mathrm{\Δx}}_3\right)-\frac{h_{2s}-h_1}{3}\·\mathrm{\Δu})$

y＝x _1s+Δx

Remain the output offset x be directly added with output thus _1s.

Can observe, difference equation is only effective to the little deviation around operation point.Operation point at this by the specified enthalpy h depending on load after vaporizer _2s=x _1sdefinition.Therefore operation point is followed the trail of by current measurement.This is realized by the variable in matrix A and B, and described matrix is obtained by the fundamental equation of linear model:

$\stackrel{\·}{x}\left(t\right)=A\left(t\right)\·x\left(t\right)+B\left(t\right)\·u\left(t\right)$

y(t)=C(t)·x(t)+D(t)·u(t),

Wherein input u (t) do not directly act under many circumstances export y (t) and thus D (t) be 0.In this way, matrix A and B are along with load or along with the enthalpy h after vaporizer 6 _2scurrent nominal value and change.This means, dynamic is mated with current load situation and on whole loading range, is followed the trail of this process thus.

Fig. 4 illustrates the schematic diagram of status adjustment.Status adjustment is linearly adjust, wherein produced difference and multiplication is joined this process by the virtual condition of process 40 compared with corresponding rated condition.If be applied to evaporator section particularly, then calculated virtual condition x (t) and predetermined rated condition x _sollt () is compared.This and below, comprise three state x in the current situation with the display of the symbol of black matrix ₁, x ₂, x ₃and Q _fas vector or the matrix of the 4th parameter or corresponding nominal parameter.Can use as coefficient and there is parameter K ₁, K ₂, K ₃feedback vector K (t).U (t) is regulating parameter, and y (t) is the output parameter of process.

In order to realize the Principles of Regulation of feedback of status, the currency of virtual condition x (t) must be known and available.But can not always measure all states now in real process.Such as can not measuring state x in current system ₂, x ₃and Q _f.Its reason is, can not determine the accurate measuring point of two states in vaporizer.The first two delay cell of model only maps the temporal dynamic property of described process.But this does not provide any conclusion to dynamic spatially, therefore can not determine the measuring point of temperature.This external state x ₂and x ₃time there is wet vapor, this additionally increases the difficulty of the enthalpy determining wet vapor.Therefore other possibilities must be found to determine these states.

State is determined or state estimation can be realized by feedback of status.By the proportion adjustment that the adjustment of feedback of status is pure.This means, state by means of only a multiplication by reverse feedback.The type of feedback may cause adjusting deviation, this means not reach predetermined rating value.In order to make these rating value be reached, the realization of quadrature components is significant.In the simple enforcement of feedback of status, the realization of integral part is solved by circuit, and the adjustment difference wherein between output value and setting value is by integrator feedback and join together in regulating parameter.

But select other possibilities in current situations, namely realization is visualizer or the interference parameter visualizer of state estimator.This state estimator comprises integral part to determine state, and remaining adjusting deviation disappears thus.In addition it has the following advantages, and namely utilizes this state estimator can estimate the interference parameter of influence process.This allows the adjustment faster to this process, because directly can see the dimension of interference parameter in estimated state.If do not have interference parameter visualizer, then interference parameter and the impact on process thereof can only be found out indirectly by the change of each state.

There are two interference parameters in current systems, these two interference parameters are considered to estimate by interference parameter visualizer.This is the ignition heat power Q be transported in vaporizer 6 on the one hand _ffluctuation, be on the other hand the enthalpy h before vaporizer 6 ₁fluctuation.But, h ₁fluctuation can be determined to the measurement of pressure and temperature by water-steam form, do not need thus to be forced to estimate.

Immeasurablel interference parameter is ignition heat power Q _ffluctuation, it has very large impact to current process.The heat value that the change of burnt main energy sources (coal, oil or rock gas) is passed through in this fluctuation causes.Thus by ignition heat power Q _fbe defined as the state Q=X of new estimation ₄.Dynamic is selected as dX ₄/ dt=0.Utilize these information can derive the state space form of the expansion for visualizer.

Visualizer is described below, because it observes described interference, so also referred to as interference observer or interference parameter visualizer.Fig. 5 illustrates the structure of interference parameter visualizer.Can find out, the model class of the evaporator section in vaporizer 6 is similar to Fig. 3, but has little change.Thus state X ₁, X ₂and X ₃state estimated by representative, wherein state X ₁=H ₂enthalpy H estimated by the output of vaporizer 6 is also described ₂, instead of real and measurable enthalpy h is described ₂.Although with capitalization, H ₂specific enthalpy is described.Estimated enthalpy H ₂with the enthalpy h passed through measured by pressure and temperature ₂whether directly compare, and will be poor, namely visualizer error e joins in observed, namely calculated process, but, but to correct the so-called visualizer of L(vectorial as with visualizer) product.This visualizer vector is four dimensional vectors, namely comprises 4 component L ₁, L ₂, L ₃and L ₄, they respectively with visualizer error-scalar multiplication.

Section state reproduced by with real concurrent process calculate dynamic segment model to carry out.Visualizer error e from the deviation between the measurement parameter of process and the corresponding value utilizing segment model to determine.Each state of segment model is respectively by with L _ithe visualizer error of weighting corrects, and makes visualizer error stablize thus.

Three delay cells 34 any one in visualizer error added corresponding correction component, its objective is and realize starting of oscillation state, namely state of equilibrium.Estimated combustion power Q and real combustion power Q _fthe 4th component X of state vector X is used as on the contrary at this ₄, and will correction component L be had accordingly ₄visualizer error e add estimated combustion power Q.

Visualizer corrects L, also referred to as feedback vector, is that visualizer error is corrected by computer at this, and namely visualizer error disappears.Visualizer can realize as nonlinear observer, because input parameter m _imeasurable.Nonlinear system directly can be rewritten as state-space representation thus.This is with the Luenberger visualizer of title-expansion or extended BHF approach device (extendedKalmanfilter-EKF)-known.With concurrent process calculate nonlinear model.But, stable feedback vector L (t) of visualizer error is obtained from linear model.This linearization is by adopting the quality of water supply stream m measured respectively _icome carry out.

Adjustment kit in the first regulating system 26 draws together linear quadratic demodulator, especially LQG regulator 30.LQG regulator is the common realization of linear quadratic (LQ) regulator and Kalman filter.LQ regulator can be so-called optimal regulator, and it is the basis of secondary quality standard.This quality standard and a kind of algorithm is utilized to carry out feedback vector K (t) of computing mode adjustment.Kalman filter is a kind of special visualizer or state estimator, wherein additionally can consider together or modeling together the measurement inexactness (measurement noises) of output and modeling inexactness (process noise).Other feedback vector L (t) of visualizer can be determined by a kind of algorithm.

Such LQG regulator is shown in Figure 6.As input to the enthalpy h measured by LQG governor assembly transmits after vaporizer 6 ₂, current quality of water supply stream m _i, enthalpy h before vaporizer 6 ₁and the specified enthalpy h after vaporizer 6 _2s, this specified enthalpy can be calculated by the rated temperature of steam 16 and pressure thereof.In addition compute matrix A is transmitted, B, A _obs, C _obs, R _regler, Q _regler, R _obsand Q _obs.

A, B, A _obs, C _obsrepresented by linearizing system and obtain, R _regler, Q _regler, R _obsand Q _obscomprise the weight factor of the controller properties (sensitivity, corrosivity) for arranging expectation.

Output is transported quality of water supply stream m _s, it is added by interference parameter m _gscalculate with the difference of status adjustment Δ m.Can observe at this, calculate interference parameter with estimated ignition heat power Q and add m _gs.Interference parameter adds m _gscontrolled in advance by Coal Quality stream in other concept, but directly calculated by estimated ignition heat power Q at this this Coal Quality stream.On the contrary, status adjustment Δ m is the result of status adjustment.

LQG regulator 30 comprises the visualizer 42 shown in Fig. 5, to the input enthalpy h that the conveying of this visualizer is measured ₁, measure output enthalpy h ₂and the confession current m measured _ias input parameter.This export-oriented this visualizer conveying feedback vector L (t) is for compensation visualizer error e.Feedback vector L (t) calculates by a solution LKR of Kalman-Riccati difference equation, transmits compute matrix A to this solution _obs, C _obs, R _obsand Q _obs.

As other assembly, LQG regulator 30 comprises for calculating rated condition X _sunit 44, this rated condition is required for feedback of status.Input enthalpy h to the input of this assembly 44 ₁with specified output enthalpy h _2s.But the not direct using state X (t) of LQG regulator 30 comes for feedback of status, but the deviation of using state and its operation point, namely with rated condition X _sthe deviation of (t).Thus as the status parameter that other will use, provide the deviation of absolute enthalpy and enthalpy rating value.Each state X _iwith its operation point X _isdeviation be 0 in operation point.If weighted sum X (t)-X _st ()=0, then do not carry out regulator intervention.Therefore state X (t) directly and rated condition X _st () is compared and is continued to use it poor.

In addition, LQG regulator 30 comprises for regulator-Riccati solution of difference equation LRR, and this solution calculates feedback vector K (t).Compute matrix A is transmitted, B, R to this solution _reglerand Q _regler.The use of feedback vector K (t) and the use of feedback vector L (t) similar.The target of L (t) compensates observer error e by multiplication and feedback, and feedback vector K (t) is multiplied with state error and for status adjustment, namely for adjustment or the regulating error for compensating LQG regulator 30 of fluctuating: from having component X ₁, X ₂and X ₃state vector X (t) with for rated condition X _sproduce the dynamic component of LQG regulator 30 in the difference of the state vector of the same three-dimensional of (t), utilize it to implement status adjustment:

K ₁(X ₁-X _1s)+K ₂(X ₂-X _2s)+K ₃(X ₃-X _3s)=Δm.

This dynamic component or status adjustment Δ m are the components of quality of water supply stream, and the component of this quality of water supply stream and the interference parameter of calculating add m _gscompare, therefore the component of this quality of water supply stream adds interference parameter and supplements.Interference parameter adds m _gsthe nominal-mass stream calculated, also referred to as basic rating value, its by estimated combustion power Q with therefrom to obtain, the business of enthalpy difference Δ h on evaporator section obtains.

This nominal-mass stream or basic rating value m _gssubtract each other with dynamic adjustments component Δ m, thus obtain specified quality of water supply stream m _s, this is the regulating parameter of the first regulating system 26.Nominal-mass stream m _sbe transmitted to the second regulating system 28, second regulating system 28 by one or more suitable parts as regulating parameter, such as raw water service pump 12 and/or valve 14 arrange nominal-mass stream m _s.

Two feedback vectors, namely visualizer correct L (t) and for regulate the calculating of the vectorial K (t) of correction be know hot dynamical state calculate known for professionals.Utilize Kalman-Riccati solution of difference equation LKR to solve filter problem for this reason, utilize regulator-Riccati solution of difference equation LRR to solve regulator problem.The solution of LQ regulator problem is undertaken by matrix-Riccati-DGL:

$\frac{\mathrm{dP}\left(t\right)}{\mathrm{dt}}=A^T\left(t\right)\·S\left(t\right)+S\left(t\right)\·A\left(t\right)-S\left(t\right)\·B\left(t\right)\·R_{\mathrm{Regler}}^{-1}\·B^T\left(t\right)\·S\left(t\right)+Q_{\mathrm{Regler}}.$

Dematrix S (t) is utilized to calculate regulator feedback matrix K (t):

$K\left(t\right)=R_{\mathrm{Regler}}^{-1}\·B^T\left(t\right)\·S\left(t\right).$

Equally also be applicable to the solution of Kalman filter problem, this problem is solved by matrix-Riccati-DGL equally:

$\frac{\mathrm{dP}\left(t\right)}{\mathrm{dt}}=A\left(t\right)\·P\left(t\right)+P\left(t\right)\·A^T\left(t\right)-P\left(t\right)\·C^T\left(t\right)\·R_{\mathrm{Obs}}^{-1}\·C\left(t\right)\·P\left(t\right)+Q_{\mathrm{Obs}}.$

Calculating observation device feedback matrix L (t) can be carried out by dematrix P (t) at this:

$L\left(t\right)=P\left(t\right)\·C^T\left(t\right)\·R_{\mathrm{Obs}}^{-1}.$

P or S is the matrix for solution matrix-Riccati-equation, and represents the intermediate parameters for determining L or K at this.

Claims (8)

1., for regulating the method being produced steam (16) by the water supply (10) in the vaporizer of apparatus of steam power plants (6), wherein state regulator (30) calculates the multiple medium state in vaporizer (6) by visualizer (42) and therefrom determines quality of water supply stream (m _s) as regulating parameter, it is characterized in that, described state regulator (30) is linear quadratic regulator,

Wherein, in order to calculate the medium state (X as status parameter _i) use the enthalpy of medium,

Wherein, use nonlinear regulated system, this nonlinear regulated system is linearized in the target offset scope of operation point.

2. method according to claim 1, is characterized in that, by quality of water supply stream (m _s) rating value be transmitted to other regulator (32) for adjustment actual quality of water supply stream (m _i).

3. method according to claim 1, is characterized in that, uses the deviation of absolute enthalpy and enthalpy rating value as status parameter.

4. method according to claim 1, is characterized in that, upgrades described operation point by adopting measured value.

5. according to the method one of the claims Suo Shu, it is characterized in that, the regulating system of state regulator comprises matrix equation, and dielectric value measured during producing steam is used to calculate described matrix equation.

6. according to the method one of the claims 1-4 Suo Shu, it is characterized in that, described visualizer (42) is the Kalman filter of linear second state Feedback Design.

7. according to the method one of the claims 1-4 Suo Shu, it is characterized in that, described visualizer (42) calculates the heat (Q entering medium in vaporizer (6) _f).

8. one kind for regulate from the vaporizer (6) of apparatus of steam power plants water supply (10) produce steam (16) equipment, there is regulating system, this regulating system comprises visualizer (42) and state regulator (30), and described state regulator prepares the multiple medium state (X being used for calculating by visualizer (42) in vaporizer (6) _i) and therefrom determine quality of water supply stream (m _s) as the regulating parameter of regulating system, it is characterized in that, described state regulator (30) is linear quadratic regulator,

Wherein, in order to calculate the medium state (X as status parameter _i) use the enthalpy of medium,

Wherein, use nonlinear regulated system, this nonlinear regulated system is linearized in the target offset scope of operation point.