WO2001067274A2

WO2001067274A2 - Method, device and computer program product for operating a technical installation

Info

Publication number: WO2001067274A2
Application number: PCT/EP2001/002479
Authority: WO
Inventors: Torsten Dreher; Siegfried Söllner
Original assignee: Voith Siemens Hydro Power Generation Gmbh & Co. Kg
Priority date: 2000-03-10
Filing date: 2001-03-05
Publication date: 2001-09-13
Also published as: DE10011607A1; AU6009601A; WO2001067274A8; DE20117191U1

Abstract

The invention relates to a method, device and computer program product for optimally operating a technical installation containing installation parts that have several adjustable installation parameters, whereby the operation of the installation is dependent on several operational quantities that can be influenced by installation parameters. Optimal values for the adjustable installation parameters (4b) are determined by means of a description (22) of the dependence of the operational quantity (13), which is to be optimized, on the adjustable installation parameters (4b), whereby secondary conditions (8) are adhered to.

Description

(-0 ω N) M H>

Cn σ cn 0 Cn cn tö φ cn

Ω tr

H

Φ

H- tr

P

P iP

H ¹

share through the existing automation systems in direct "interventions in the technical system ^" .

^'' Many parts of the technical system can usually be set using system parameters to perform their tasks. The current operating point of each system part results from the current values of the associated system parameters. The operating status of the technical system results from the operating points of the system parts; in general, several or even all system parts of the technical system contribute to the operating state of the technical system.

For example, there are several turbine generator units in a technical system for generating electrical energy, each of which works at a specific operating point. At their respective operating point, each turbine generator unit generally has a large number of operating variables, such as the efficiency, which depends on the respective operating point.

If you now consider the overall efficiency of the technical system, it depends on the efficiency of each subsystem at the respective operating points. A change in an operating point of a system part by means of the respective associated system parameters thus leads to a change in the operating size of the technical system.

Tasks are often given to the operating personnel of a technical system to set an operating size of the technical system to a certain value. ^' Since the operating size of the technical system cannot be set directly, the task is solved by setting the system parts using the associated system parameters. In general, there are several combinations of system parameters that fulfill the task, but these are general with regard to other criteria LM N>F> *

(_π O σt 0 Ln O cn

s; ιp Ft F ¹ ro NF ¹ ; P * 5 ö H a α g p. er α tö Ft OF ¹ B- g TJ w er Z rt gat? 3 a

F- Φ P P> H F- P F- o-. «PP P> ro - P. ro P ro a tr F- rt. rt C Φ O φ F- rt F- rt P ιP cn Φ Ω ai HP x cn tr Hi cn tr rt φ Ω n F ¹ P rti Hi a tr Ω F ¹ Cn N Ω "P Φ Φ ro tr a Hi Ω v § & ι:: cn ro H α a tr Ω PP ^ ro ro tr P Ω ro tr

C p ro F- ö H Φ ro rt P "υι φ X a H ro F- F- M Φ tr ι F- tr PP F- P ιq tr P rt ap rt ιP et £ £ P p ≤ 3 a F- v H ro P ro cn rt tr F- CD ro P ι rt- Φ cn rt P> ro H F- Φ P s: er 1- »a tτ <α P Φ ^ N cn a PL cn o N Ω 1- 1> H F- öa ≤: rt na F- cn J ιP O: P ro F- aa rt P Ω tr a t-> ip tr o CΛ tr et P F-> ro P F- cn IT Hi ι ι- <--1 Φ cn cn H a ι rt cn tr P o ιP Ω Hi ro cn ag L Φ rt tö r cn P> H Ω ro O P. ιp PF ¹ cn tr ftv. Ro rt cn F- I Φ p Φ tr 0: a X P- tu rt- ¹ α tr F- H P- 0 »Pi Φ a tö P s: p: N rt a F- K Ω et PF ¹ tr P- F- ro a ro p rt P. ro α P CD tr ι φ a ro ro PPN a P 0 tr rt p P rt ro 3 F ¹ cn ro Φ Hi Φ cn ro Φ P. ι rt rt rtl F ¹ P LP cn X p. 'F- ro Ξ cn p P! P> Ω aa rt P. n tι Hi LO a α F- rt ιp rt- F- er F- & a ro rt- ro

P ro a ro cn a PP ι tr et Φ F ^ OF »Φ φ Φ F ¹ ro ιP ro Ω F- ro rt tr PH Hl> P ≤: a 0 Φ h cn F- cn F- & cn cn a HP tr Φ a P tr P P- cn a rt

Ω cn ao F- ιP cn a F- 0 Ω Ω Ω ¹ F- cn JP F- P. Ω ro ip F- φ F- tr ιP ^' a ro F ¹ P. ti ro ro H a- V <tr PH F-> £ φ et Ω ro P tr ro ro F- 0 LP

H rt- a * P ro rt a α s: tr i F- φ ro Φ ιP rt F ¹ F "ro P rt • tr F- o ro rt cn ro F- ro ro Φ P hi φ cn F- Hi cn tr PB P- rt- ^ ro ha α • θ: r 3 • ro

Φ EP a rt Ω 3 H a a α L ≤: a Ω r c rt- v O ro p g Φ rt ro 0 Φ * F- s: a- ro F- ro P a s P cn rt P F- F- P rt ö

Φ tr rt tr- Ω rt rt Hl tJ ro

F- p CΛ PP P> H <J cn aa P ro H F- & ι HH ti rt- P ro ro t 01 ro et LP P ro rt ι Ω ιP & 3 Hi ro Ω <3 _^ »Hl F- rt H Hi ro P- φ ro F * p. a ^ *. F- rt Φ P cn F- P p P. tr ro a rt H a- ro ro s: Ω cn ^ ϊ- rt ca F- P F- F "tr a rt

Φ ro HP ⁾ ro ιp i H a 0 P tr Ω P ^" F- ¹ ro Φ F- Φ ro tö rt P. ro cn ro tö rt rt- rt F- g ^* F- rt - Hi Hi tr P cn tr P F- H F- «3 F- ro - ^ P * a» ro 3 F- ro Hi Hi Ω tr ro rt ⁾ Ξ " _^ P. P ip Ω rt- a H a cn rt cn P, p. P ιp rt- rt p: ro rr ro F- F ¹ tr ro φ Hi cn tr ro F- K Φ O rt P. cn ö tr Φ ro F- ro g ro rt a ro cn Ω F- HH cn ro rt ιP Φ a <rt F- F- F- P ro rt F- φ F- Cd pi Ω F- p cn tr Ω ro P. i cn F ¹ H Φ Φ P ιp rt Φ ro ro cn F- a 3 t P p- cn Φ ro P "a Φ P <F- P! 25 HH tr c ro tr cn cn ro T aJ P rt pp rt Φ H a rt cn a PO £ Ω P. PH Hl. F- F ¹ rt P cn ro tö ^' J & Φ F- F-

P ⁾ F- P Hi P ro £ 0. P F- tr φ rt PP Φ F- Φ F- cn cn. ro, F- F- LP rt φ rt cn N o et PH TJ ro PH Φ cn N Hi tr a Ω F- F- • rt- α ro P ro cn rt ro

Q Ω s: to rt N Ά rt F- a tr rt Φ r + H JT O cn rt P ro F- F 'ιp Hi 0 rt

Φ P Φ P. ro cn 2: F- cn P- N W c a £ φ α F- P P ro cn φ 0 a S3 F- p φ H rt ro P 3 ro a ro 2; Ω P P ro a P P Φ ro P Hi fi ≤: t H P> cn Pi 0 tr & Ξ rt P F- ro

P r Hl <1 HH p: H rt et ιq J ^• Φ J α F ^J rt a

N F- F ¹ ro &> ro H F- PX 0? R N Λ Hi rt- F- rt F * ro F- P ro wpg F- φ F- F- PP cn Φ Hi Hi PP a tn P rt- cn ro OPP rt cn rt P. P - ro er Ω PP> rt rt- rt- H: rt a PP hi ro O aa ιP cn ro P rt- φ a O: o F 'tr φ p *> - • ro HF ¹ n £ a F- Q t → Φ P. Φ 0 .3 ro rt P a J ⁾ ≥; P. ro H 3 F- rt ro ≤: F- £ PF ¹ O F- ro PO: a PP - ^' ro cn a et ιP o Φ p cn ro Ω H cn Φ <l Ω Hi tr 03 PJ rt »rt cn CΛ rt- P φ r - t et • LP ro

F- φ rt- H φ a tr F- F ¹ Φ tr Φ Φ a ιP rt tr P P. ro ro F "H. Ro ro ro rt aas:! F- ro ro a Ω H Φ cn HP ro F- Φ a Φ F- F- rt <-r Ω P

F- x ro J! & rt PP tr 0: P ^* tr a rt tf a 3 tr F- F- v rt a F> F ^ O rt o ^ Φ φ Φ tr 3 F- cn Φ Φ φ O: er Φ rt- 5 a ^" N φ. F- a rt H) cn ar • rt ro P- P s: s: a cn ιp F- Φ P Φ ro JP rt rt in ro F- ro F- tö ro F- Hi P £ F- F - <cn P Φ ar F- F- t φ φ 3 rt aarn ro p N ι rt p PH Ξ Ω Φ ro rt ω P aatr ι P tr Φ Φ rt Ω ω F- N cn Ω rt

Ϊ X p. P- Φ rt ro & HHN rt PN φ P rt- F ¹ cn p- Φ rt PP iP t 3

"tö Φ Φ 3 cn H r P Φ φ cn P ι ι tr rt P rt rt- F- a cn rt Φ r F- ι_t. F- HP Ω Φ aa F- i Φ 0 P 3 F- P ro φ LP rt- O: P Φ a

Φ rt! * R P Φ α tr Φ cn vp cn F "P- er P- Hi Hi O: Hi cn H, H cn Φ ro fco Ω F-

P P- cn 3 i ^j a P> cn cn rt ro Φ φ Hi rt ιP ιp Ω cn cn ro ι H ro tr 3 o ro pi Φ tr Hi CΛ φ 3 P • H Φ a cn ro 1 P tr 1 Φ F ¹ 1 P

Ω p PH Hi ro rt cn 1 P a H. • a tr φ 1 ro ^J tr 1 1 0 F- l α 1 • Φ a, a ro - ro H 1 a rt

there are no indications of the specific measures that could be used to do this. It is only assumed that such methods for optimizing an operating size of the technical system are known. The comparison between the need for the implementation of the optimization effort and ultimately resulting economic benefit is the core idea ^'of ^the' above-mentioned publication. To get the full Mutzen from the disclosed in the document process, it is necessary to effec- possible tive process for the optimization of a company size, such as the efficiency of a technical system, which can then be used for the aforementioned comparison.

The invention is therefore based on the object of making one or more operating variables of a technical system intelligently adjustable with a plurality of system parts and of optimally setting them by means of the system parameters of the system parts, whereby secondary conditions must not be violated.

According to the invention, the method of the type mentioned initially consists of the following steps:

1. At least one of the operating sizes of the technical system is selected for the optimization.

2. A mathematical and / or measurement-based description of the dependency of the operating size of the technical system to be optimized on the adjustable system parameters and marginal parameters that necessarily result from the operation is created.

3. The current system parameters and the current boundary parameters are recorded.

4. Additional conditions for the adjustable system parameters are specified. 5. For the current operating state, captured by the current system parameters and boundary parameters, under Compliance with the secondary conditions optimizes the system parameters. 6. The system parts are set using the optimized system parameters. ^'

In this context, “mathematically and / or measurement-based description” means in particular analytical mathematical formulas or characteristic curve fields consisting of discrete data points obtained at support points.

The method of the type according to the invention does not require any change in the hardware of the components of the system parts; Such changes may also contribute to an improved mode of operation under certain circumstances, but ■ are advantageously made during the design of the system and no longer during operation, which would be very expensive. In the method according to the invention, the system parts of the technical system are intelligently set so that optimized operation results and auxiliary conditions are not violated.

Such secondary conditions can be, for example: - a plant ^• for the generation of electrical energy is required to provide a fixed total output - which is made up of the sum of the individual performances of the plant parts - and each plant part only in one permissible operating range may be operated. Then, for example, the production costs of the technical system or the efficiency have to be optimized, whereby the previously mentioned secondary conditions have to be observed. The above. Operating parameters Production costs and efficiency of the technical system are only to be understood as examples - the method according to the invention is also suitable for optimizing a large number of other operating parameters of a technical system. OJ LO NO) t-> F ¹

Cn O Cπ o Cπ O n

P t Q. cn Φ F- Fi r 3 t

& t P Φ ro ro rt o P F- F- rt et rt PP s: F s ^* cn 3 r P- tu ro

Ω ro ¹ tr rt- g tr tö LP P. lP O op: P ro F ¹ F- ro TJ r

F ¹ a rt Φ a rtl rt rtl F ¹ SV rt et rt F g 3 PPP rt 0 rt- 3 a ö φ Φ 0 P F-. P Φ ro PP ro P r * a - cn ag et Ω F "CΛ ro

P rt P Ω LP ro r tr Ω rt rt- LP tr φ s ro P rt ro P LP rt F- LP o F- F- ro P rt ro P n LP O ro o

F- ι rt- rt ιP F- o rt- Ω 0 rt 3 a rt a rt- t- 8 F- LP Ω ro a rt rt- rt P φ rt- S et ro ro «q P a tr F ¹ ro ro P ro LP rt ro ta rt- ^■ α a ro Ω ro r X rt- a ro F- F- ro θ:

F- er P φ ro p: cn P er - rt ^' tr NP ro ro PD r F ¹ er tr a F ¹ p: ro ro rt co U) P p ιp P cn TJ Hi rt a rt- ro a ro ro cn Φ Φ rt rt a ro ro P cn ro J F- tr cn rt ro Ω cn H ro <q P 0- 3 P ιP a ro a P 0 F- F- F- a rt- rt- P rt a LP N a P Ω F ¹ ro rt- a F- tr rt- t rt tr rt F- P F- • rt P rt F "rt rt ro or cn rt Φ P rt p t- rt ro ro rt ro ro g F - r - r Ω s O: P rt rt rt ιP cn F ¹ Ω tr O: F- l P LP rt- tα £ cn 01 Φ ro rt rt ap F ¹ - F * tr φ to 3 rt ro F-! * T Φ S tr aa P BD ιP φ ro ro ro ro Hi ro tö g Φ J rt P 3 cn

" _* at rt F ¹ P Ϊ cn ro φ P ι Φ <J 3 er rt PP er ro ro t

Cn rt- F ¹ ro F- Φ cn F- rt g 0 g er P ro a rt rt. a Φ ro F ^ p: cn rt rt a ro rt- t tu rt ro Ω P ro F-

TJ a ro X ro rt rt O cn P ro P rt et P tP N F- QJ rt ro Φ ro rt et F- I F- rt et

P Φ rt oa to tr LP et tr Φ N rt rtl rt l Φ P φ 3 rt a et F- rt P 3 ~ a cn cn r ro P φ ro F- P ro 0- rt- rt- ro cn tr. φ 3 P. rt- rt ro rt- LP 1 PN F- ro g cn o F "P cn F ¹ a rt rt ro ro a <rt cn rt- ro PN ro rt- tr ro ro ro 0 rt ro rt s P. F- Ω a rt- P rt- F ¹ cn o P Ω er TJ Φ P n fD F- X a ro cn cr a P rt- F-

Φ Φ F "tr ro P ro ro rt- TJ. Cn rt- cn tr CΛ P rt P F- rt Λ ro P T3 er N cn rt- P <3 ro P

F ¹ rt a P rt φ PP tr P ro et vP ro OPN rt rt rt- cn rt ro cn PN ro cn OP "ro er P (S LP Φ F- rt- ro P F- F- ro Ω rt F- PPPP ιp rt l cn P rt- rt- P Ω: CΛ

Φ LP P. 3 ro F- a φ ro a rt 3 cn tr ιp cn cn 3 tr P- P F- ro N Φ cn rt rt cn F- ¹ fD tr o cn Φ P a tr ro F- rt er F - et a ro Ω et ro n ro et tr F- P F- OO: P rt ro F "a ro aa

Ω a P. P rt- P 3 a tt tr F- P ro P F- a tr P rt- ro rt F- rt a rt aa &> PPF ¹ ro a ro ro tr rt- P * - ro a ro ro ιp rt rt F "cn - ro P ro a O ro Φ P ro a P ro ro rt. cn rt rt ro rt rt F- P g rt p rt- Φ ro rt Ω P rt rt- a MPF ¹ a F- a

Φ F- Ω P F "a a a a P tr P r ro Φ X <rt o F- a tr ro

F- tr ro N F "ö F- ro 3

P aa • Φ P g 1 tr ro Ω a F- F- rt LP a TJ 3 a NN 0 ro rt ro er cn F "o PP Φ F- F- rt ro LP P cn ro cn tr rt ro ro Φ P rt ro 0 P Hi Hi rtl tr CΛ ^'

Φ <F "α LP rt aa & a F ^J P rt ro PO ¹ F- ft F- Λ rt cn rt F- rt a N Φ ro Φ ro rt F- Φ ro F <Φ PP ro F- F- N Ω O Ω 3 F- and 3 P ro a ro

P rt ro rt rt P F- ro rt cn P ö ro & F- LP cn F- rt cn ro cn P tr rt tr P a rt- ro 0- er F- a rt

P ιP P F- p- TJ φ a rt- LP ro rt F "a ro LP rt F- Ω rt ro tr 0 a F" a ro F ¹ cn ro rt P ro

F ^J ro rt a PPP tr Hi rt- ro φ et P Φ φ cn tr F "F <P Φ 3 ro Ω F ¹ cn a cn o cn lP a F ¹ rt cn ro ro a rt w ιP ro α rt- et ro CΛ rt- lP rt P F- tr ro ro a ω rt P ro rt P LP F ¹ Ω h ~> TJ rt- rt ro O rt rt fD a TJ ro 3 rt a cn appa

P.. Ω cn a Φ 3 rt tr et P ro rtl p. r § P cn ro

P t 0 H α ro a 0: a φ P rt ro F- cn rt

F- a tr rt- ro a ro o-. ro a ro rt tr F- O ro ro Ω F ¹ rt- NP rt- a ro _* rt Cn cn ro ro rt cn ro rt- ro cn * _* rt- &> rt F- G) cn PJ rt F- w rt ro & F- o F * ro F ¹ Ω F- X rt tö 3 rt-

? T rt ro F "CΛ ro φ CΛ a 3 P rt Ω et cn F ¹ P p. F- cn tö F * F) rtl tr P ^' rt- F ¹ ro P = ro rt Y-' • ro X rt N Ω F- φ rt S F- 3 tr ro PF ¹ ro Ω ro Φ ro 3 i ro a ro ro a & er φ XP rt- P φ Φ P tr cn rt- O: a 3 ro ro a LP rt tr cn et cn rt p rt- ro cn rr F- S * ro rt er F- rt ro! ro S> ιP rt- ta P ^' rt- ro p: ro rt a P ro a LP

Φ rt rt a trt P-- Φ aa rt rt ro ro rt rt ro s F ¹ a rt P rt- MF ¹ a cn g rt- 3 p rt α PP φ co ro <ro as; rt ro P Ω P ro F ¹ 0 ro ro F- F ¹ rt- 0 ro lP TJ <j O:

^ J ^• • - ro s er a Ω cn ro rt- α F- rt rt ar N a rt ro rt tt tr P lP ro F ¹ ro ro ro Ö3

PF ¹ ro o tr rt & ro P ro rt ro P. ^ ιp rt a OO £ Λ CΛ ro P ap rt rt ro rt 3 F "rt rt 3 rt rtl F ¹ a rt P- X α 3 ro ro F- rt N rt- 3 ιp P cn rtl P

<i φ ro rt- 3 Φ & p: P s! a ro PN F- F ¹ F- ro a & a P ro ro J ro cn g * 1 0 P ro tr o F- F- F »LP LP or rt- P. rt P p φ ro • _^ Ω F- F-. ro cn * F- rt P cn P tr ro rt Si ro a cn r PP ro rt? ü ro tr- F- tr - P rt- a p. cn 3 a. rt et F ¹ a F-. ιp 3 P rt F- a ro ro F- ro rt Φ P a rt- P Ω ro ro LP ≥ia F- rt ro P rt- P ro ro 3 ro cn ro φ F "Φ P

P rt rt P F- Φ LP rtl ro a tr P Φ P φ ro a. O rt rt ro a P a rt- F- F- a F- P ro cn ro rt φ cn ro s P _* 0- a rt- rt tr T £ J- • - tr a

F- S £ p: a TJ ro F- a a fD a Ω J Cfl a cn rt α Φ P- ö rt F- O a? R rt cn - tr P rt ro α cn r P.

1 ö o ro P CΛ 1 tr ro a ro ro P F- rt a 3 ro PP rt- 1 rt o rt- ro 3 F- F- a 1 rt Ω F ¹ rt tr F- 3 cn l cn ro ro rt a P ro ro. TJ ro rt F ¹ ro F- P rt- rt- ro Φ aa F- P LP F- rt Ω cn ro 1 ro ro a 1 ro ro CΛ F ¹ a F- tr P a cn rt rt ι 1

, - ~ J

already existing automation system of the technical system can be implemented if there is data connection between this and the other automation systems of the technical system and thus there is access to the system parameters and operating parameters of the other system parts.

Furthermore, the invention leads to a computer program product which can be loaded directly into the internal memory of a digital computer and comprises software code sections with which the steps of the method according to the invention are carried out when the product is running on a computer.

Three embodiments of the invention are explained with reference to the accompanying drawings. It shows:

1 shows an industrial plant connected with a computing unit for implementing the method according to the invention, FIG 2 is an arithmetic unit for carrying out the method according to the invention with an interpolator for interpolating the current values of the operating state, and ^'3 shows a technical facility connected to multiple parts of the plant with a computing unit for carrying out he method according to the invention with several processing units for interpolation and for carrying out various mathematical operations.

1 shows a technical system 1 with a number of system parts 3, 5, 7,... N, which are autonomous insofar as each

^'' Plant section only requires input of plant parameters, which then control the plant section (e.g. via its own computer).

The system is connected to a computing unit 20 for carrying out the method according to the invention. The computing unit 20 contains an optimization stage 24, the one

mated system parameters 4b; the plant parts are optimized by means of these system parameters 4b ^'then set.

3 shows a technical installation 1 with a plurality of installation parts 3, 5, 7,... N and another embodiment 20 'of the computing unit. Fig. 1 shown. The computing unit 20 'contains an input unit EE, for example a keyboard, with the aid of which the discrete values 23, which are preferably determined by measurement at fixed support points, for technical description of the behavior of the operating variable of the technical system to be optimized are recorded as a so-called characteristic field. A first interpolator INT1 determines from the discrete values 23 of the characteristic field a preferably section-wise analytical mathematical representation of each characteristic curve of the characteristic field described by discrete values. A section is delimited by a first support point and a second support point immediately following it.

Commonly known methods for interpolation are linear interpolation, interpolation using polynomials or rational functions and interpolation using so-called splines. Interpolation by means of splines is preferably used for the interpolator INT1. The interpolation between the discrete values 23 obtained at interpolation points is carried out in sections by low-degree polynomials. A polynomial is assigned to each section. A so-called spline function consists of the polynomials assigned to the sections; the polynomials are all of the same degree. Third-degree polynomials are used with particular advantage and thus obtain a so-called cubic spline function. Cubic spline functions have a smooth 'first derivative in the mathematical sense and therefore a continuous second derivative. Both derivations are available analytically, since the derivation of polynomials - and in particular of third degree polynomials - is known. When using spline functions are third degree consequently, to determine four spline coefficients for each section which represent the coefficients of the associated polynomial. The following conditions are required to determine the spline coefficients:

1-. The spline function must assume the discrete values specified by the technical description at the support points and be constant.

2. The first derivative of the spline function must be constant in the support points.

3. The second derivative of the spline function must also be constant in the support points.

In addition, the second derivative preferably takes the value zero at the outer support points, that is to say at those support points which only belong to one section - the so-called outer support points. Algorithms for determining the spline coefficients in sections under the aforementioned conditions are known. For each characteristic curve of the characteristic field - which is described by the discrete values 23 - which belongs to a fixed value of a • boundary parameter, the first interpolator INT1 determines using cubic spline functions for interpolation for each section of the by the discrete values 23 characteristic field described the spline coefficients; for each of these sections, the interpolator INT1 thus determines four coefficients when using cubic spline functions. The first interpolator INT1 interpolates - that is to say via a first variable - specifically via the system parameters 4a - by analytically determining spline coefficients 32 for each section of a characteristic curve of the characteristic field, which belongs to a fixed value of an edge parameter. These coefficients 32 are advantageously assigned to the corresponding sections and stored in a memory SP. The previously mentioned steps of acquiring the characteristic field and interpolating over a first quantity need only be carried out offline once; co co r ro F "

Cπ o Cn o Cπ Cn

The value of an edge parameter is smaller than the current value of an edge parameter for which the characteristic is to be determined, and the second fixed value of an edge parameter is to be greater than the current value of an edge parameter for which the -interpolated characteristic is to be determined. The educational requirement is then: ■

c (p) = ai ^• Cι (b _λ ^• p) + a ₂ • c ₂ (b ₂ \ p)

The two characteristic curves adjacent to the characteristic to be interpolated, which belong to fixed values of an edge parameter, are therefore superimposed. The curve shape of the characteristic curve c (p) to be interpolated should be more similar to the curve shape of cι (p), the closer the current value of a boundary parameter belonging to the characteristic to be interpolated is to the first fixed value of the corresponding boundary parameter; The same applies to c ₂ (-p). The coefficients ai and a _{2 should} be chosen accordingly:

The coefficient a _{2 is} determined as the quotient from which

Difference between the current value of an edge parameter and the first fixed value of an edge parameter and the difference between the second fixed value of an edge parameter and the first fixed value of an edge parameter; the coefficient a is then chosen as follows:

The coefficients bi and b ₂ are advantageously calculated as follows: the coefficient i is determined as the quotient from the first fixed value of a boundary parameter and the current value of a boundary parameter and the second coefficient b _{2 is} determined as the quotient from the second fixed value of a boundary parameter and the current value of a boundary parameter. A test of the interpolator INT2, which is intended to work as described above, shows that this maps the characteristic curves belonging to a fixed value of a boundary parameter to them themselves if the current value of a boundary parameter corresponds to a fixed value of a boundary parameter ,

The output of the interpolator INT2, which provides an interpolated characteristic for a current value 15 of an edge parameter, is routed in parallel to two processing stages VS1 and VS2. The processing stage VS1 uses this to calculate the current value 40 of the farm size selected for the optimization. The processing stage VS2 calculates the current value 42 of the gradient of the operating variable selected for the optimization; For this purpose, both processing stages VS1 and VS2 are supplied with the current values 4a of the system parameters. The current value 40 of the operating variable selected for the optimization is transferred to a display unit BI, for example an output field of a computer screen, where the behavior of the operating variable to be optimized then during the operation of the. technical system 1 can be monitored. The current value 40 of the operating variable selected for the optimization or the current value 42 of the gradient of the operating variable selected for the optimization are made available to an optimization stage OFT.

The optimization stage OPT processes an optimization algorithm, which is advantageously a method of so-called sequential quadratic programming - abbreviated SQP. SQP methods are currently considered to be the most effective for solving general non-linear optimization tasks. Details of such a method of sequential quadratic programming can be found, for example, in Markos Papageorgiou: "Optimization: Static, dynamic, stochastic methods for use", Oldenbσurg, Munich, Vienna, 1996. The optimization level OFT also receives the secondary conditions 8 to be observed during the optimization, as well as start values 45 for the system parameters with which the optimization process is initialized.

The quality of the optimum for the optimized system parameters 4b determined by the optimization stage OPT may depend on the choice of the starting values 45. The optimization process, which is implemented in the optimization stage OPT, is advantageously started several times, with different start values 45 being used in each of these runs of the optimization process. The optimized system parameters 4b belonging to the best optimization result are then transferred to the technical system 1; the system parts 3, 5, 7, ... n are to be set using these optimized system parameters 4b.

The technical system 1 shown in FIG 3 and the computing. Unit 20 'can be used in a particularly advantageous manner as a device for optimizing the overall efficiency of a hydropower plant that contains several machine sets. This power plant can be required, for example, that it provides a certain total electrical output for a network, that each machine set only works in a certain permissible operating range, and that the overall efficiency of the hydropower plant is optimal. In this case, the size of the hydropower plant to be optimized is the overall efficiency. The secondary conditions consist of the condition for the total electrical output to be delivered and the conditions for the operating areas in which the individual machine ^• sets should work. If the permissible operating range of at least one machine set consists of several non-overlapping partial areas and / or operating points, this cannot be used when using, for example, an SQP method Supplementary condition in the form of exactly one equation and / or inequality.

In this case, the solution of the problem can be to go through the Optimierύngsverfahren several times, wherein each one of the admissible partial regions or operating points is formulated as just a constraint on these ^'runs successively; The best optimization result in comparison provides the optimized system parameters (here: power setpoints to be set for the machine sets). This combinatorial procedure may result in an increased need for retirement time; One starting point for improving the treatment of the problem of secondary conditions, which are composed of several permissible sub-areas and / or points and thus cannot be directly processed in many known optimization methods, is the use of strategies of so-called genetic algorithms, which Literature are known.

An essential boundary parameter of a hydropower plant is the currently available usable head of the water. As a rule, a characteristic curve field is available for each machine, with a fixed value for the drop height of the water being assigned to each characteristic curve in this characteristic curve field. Each of these characteristics describes the course of the efficiency of the respective machine set over the electrical power output by this machine set. The first interpolator interpolates via these outputs, the second interpolator determines the interpolated characteristic curves of the machine sets based on the current value of the water head. The first

Processing level forms the current value of the overall efficiency of the hydropower plant and urges this value to be displayed; the second processing stage calculates the current value of the gradient of the overall efficiency. For this purpose, both processing stages use the current values of the outputs set on the machine sets. The optimization level optimizes the current value of the overall grades and calculates optimized performance values for the machine sets, with which these can then be set to achieve optimal overall efficiency. In this way, an operating variable (here: the overall efficiency) is optimally adapted to the boundary parameters and secondary conditions that cannot be influenced. '

Claims

claims

1. Method for the optimized operation of a system part (3, 5, 7, ..., n) with a technical system (1) containing several adjustable system parameters (4b),

• the operation of the system being dependent on several operating variables (13) which can be influenced by system parameters (4b), with the following steps:

a) at least one of the operating variables (13) of the technical system (1) is selected for the optimization; b) a mathematical and / or measurement-based description (22) of the dependency of the size of the operation (13) to be optimized - the technical system (1) on the adjustable system parameters (4b) and the parameters (15) resulting from the operation created; c) the current system parameters (4a) and the current boundary parameters (15) are recorded; d) constraints (8) for the adjustable system parameters (4b) are set; e) for the current operating state detected by the current system parameters (4a) and boundary parameters (15), the system parameters (4a) are optimized while observing the secondary conditions (8); and f) the system parts (3, 5, 7, ..., n) are set by means of the optimized system parameters (4b).

2. The method according to claim 1, characterized in that the dependency of the operating size (13) of the technical system (1) to be optimized on the adjustable system parameters (4b) and boundary parameters (15) is described by table-like discrete values (23).

3. The method according to claim 2, characterized in that interpolation calculates a current value (13) of the at least one operating variable (13) from the discrete values describing a current • operating state, this current value (13) of the ^' at least one ^' operating variable (13) (4b) are calculated and optimized from ^'the optimized system parameter;

4.Device for the optimized operation of a system part (3, 5, 7, ..., n) with a number of adjustable system parameters (4b) containing technical system (1), the operation of which depends on a number of operating parameters (13b) which can be influenced by system parameters (4b) is, wherein a) in the computing unit (20) at least one of the

Operating parameters (13) of the technical system (1) are defined for the optimization; b) in DER computing unit (20) comprises a mathematical and / or measurement technology based description (22) of Abhäng- accuracy of the necessarily to be optimized operating parameter (13) of the technical installation (1) of the adjustable plant ^'parameters (4b) and marginal parameters (15) resulting from the operation are contained; c) the current system parameters (4a) and the current boundary parameters (15) are recorded in the computing unit (20); d) the secondary conditions (8) for the adjustable system parameters are recorded in the computing unit (20); e) the optimization stage (24) contained in the computing unit (20) is trained to add the boundary parameters (4a) for the current operating state, ascertained by the current system parameters (4a) and boundary parameters (15), while observing the secondary conditions (8) optimize; and f) the system parts (3, 5, 7, ..., n) are adjustable by means of the optimized system parameters (4b). Computer program product which can be loaded directly into the internal memory of a digital computer and comprises software code sections with which the steps according to claim 1 are carried out when the product is running on a computer.