Embodiment
For making the object of the invention, technical scheme and advantage clearer, will combine accompanying drawing that the present invention is done to describe in detail further below.
Fig. 1 is the process flow diagram that the transport function of system object parameter model of the present invention is confirmed method, comprising:
S101: the transport function of predetermined system image parameter model comprises appearance time lag links such as n rank;
S102: select preset several identification Frequency points, obtain the amplitude transfer coefficient of each identification Frequency point and the phase transfer value of said each identification Frequency point;
S103: the amplitude transfer coefficient according to the identification Frequency point calculates said inertia exponent number n and the inertia constant of holding link that wait;
S104: the proportional gain and the time lag constant that calculate said transport function according to current inertia exponent number and current inertia constant.
Because thermal power plant's heat power engineering system image parameter model also has pure time-delay characteristics except having characteristics such as big inertial time sex change.So, for satisfying application of practical project, need to confirm that system object parameter model model is appearance time lag links such as n rank, be specially:
In the middle of prior art, the time domain of employing is debated knowledge need use all continuous frequency spectrum resources of thermal object, and result's influence that spectrum environment is debated knowledge to time domain is very big, disturbs owing to exist, and therefore time domain is debated the poor effect that knowledge obtains.If can identify the continuous full spectrum of frequencies characteristic of object, also can identify the object model transport function according to continuous full spectrum of frequencies characteristic fully.But see then very difficulty and unrealistic from engineering practice.The present invention selects preset several identification Frequency points; Only be that this limited several frequency characteristic of unifrequency point is debated knowledge; The frequency characteristic data that obtains these unifrequency points is very easy to; And use BPF. that these unifrequency points are debated the knowledge frequency signal and carry out Filtering Processing, therefore have stronger anti-interference again.This also is to be made up of various frequency contents because of undesired signal, and " bandpass filtering " maximum characteristics are can most unwanted frequency compositions be discharged outside, so the Points Identification for Process Frequency Response sharpest edges of unifrequency point also are stronger anti-interference.Preferably, the present invention only need select the frequency characteristic data of 3 limited identification Frequency points, just can debate knowledge fully to appearance Object with Time Delay Model Transfer functions such as high-orders.Preferably, choosing 3 identification Frequency points is respectively:
![Figure 687597DEST_PATH_IMAGE002](https://patentimages.storage.googleapis.com/37/89/12/8c41498d395b36/687597DEST_PATH_IMAGE002.png)
,
,
; Obtain the amplitude transfer coefficient of each identification Frequency point:;
,
; The phase transfer value that obtains said each identification Frequency point is:
;
,
.
Fig. 2 is the first embodiment process flow diagram that the transport function of system object parameter model of the present invention is confirmed method, and compared to Figure 1, Fig. 2 is the concrete refinement step of step S103.
S201: confirm first functional relation between inertia exponent number n and the inertia constant T according to first group of amplitude transfer coefficient;
S202: confirm second functional relation between inertia exponent number n and the inertia constant T according to second group of amplitude transfer coefficient;
S203: calculate inertia exponent number n and inertia constant T according to said first functional relation and said second functional relation, wherein, the said first functional relation non-equivalence is in second functional relation.
The object model transport function is (1) formula:
(1)
Ask for the frequency domain amplitude characteristic function and obtain (2) formula:
Amplitude transfer coefficient
and frequency
thereof for people (2) formula, are obtained (3) formula:
Amplitude transfer coefficient
and frequency
thereof for people (2) formula, are obtained (4) formula:
Amplitude transfer coefficient
and frequency
thereof for people (2) formula, are obtained (5) formula:
Arrangement (4) formula obtains (6) formula with (5) formula:
(6)
(6) formula is obtained first functional relation between inertia exponent number n and the inertia constant T:
(7)
Arrangement (3) formula and (4) formula obtain second functional relation between inertia exponent number n and the inertia constant T:
Can learn according to said first functional relation (7) and said second functional relation (8); Have only n and two unknown numbers of T in these two function formulas; And first functional relation (7) non-equivalence is in said second functional relation (8); So, be the value that can calculate n and T according to the equation principle of mathematics.Particularly, be the first embodiment synoptic diagram that the transport function of system object parameter model of the present invention is confirmed method like Fig. 3.Wherein, horizontal ordinate is represented inertia exponent number n; Ordinate is represented inertia time system T; Curve 1 expression first functional relation; Curve 2 expressions second functional relation.So, the point of crossing Sn of two curves is just by being asked.
In addition, said first functional relation (7) is one embodiment of the present invention with said second functional relation (8), and the present invention does not get rid of and can also obtain have the function formula that realizes identical function through other similar mapping mode.For example, above-mentioned first functional relation (7) is that amplitude transfer coefficient through
and
obtains.Certainly; Under the prerequisite that guarantees first functional relation and the second functional relation non-equivalence, can obtain through the amplitude transfer coefficient of
and
.
The present invention can only select the frequency characteristic data of 3 limited identification Frequency points, can choose the Frequency point more than 3, selects several Frequency point data wherein to obtain first funtcional relationship and second functional relation.For example, can choose 4 identification Frequency points is respectively:
,
,
,
.Then, using the
,
to obtain the amplitude transmission coefficient of the first function formula; using
,
amplitude transfer coefficient for the second function formula.
The proportional gain of calculating said transport function according to the amplitude transfer coefficient of identification Frequency point, current inertia exponent number and current inertia constant.Preferably, arrangement (4) formula obtains gain coefficient formula (9) formula:
Certainly, also can ask for proportional gain K through the formula that concerns of (3) formula or (5) or other Frequency point.
The time lag constant that calculates said transport function according to the phase transfer value of identification Frequency point, current inertia exponent number and current inertia constant.
Object is worth theoretical formula in
frequency plot transmission and is (10) formula:
(10)
For people (10) formula, and arrangement obtains the calculating formula of time lag constant with phase transfer value
and frequency
thereof:
Certainly, the above-mentioned formula of phase transfer value substitution (11) that also can be through
or
or other Frequency point is to calculate the time lag constant.
Fig. 4 is the second embodiment process flow diagram that the transport function of system object parameter model of the present invention is confirmed method; Compare with first embodiment of Fig. 2, Fig. 3; First embodiment illustrates on principle can obtain inertia exponent number n and inertia constant T comparatively exactly through calculation procedure; Second embodiment is then from practical application; Obtain inertia exponent number n and inertia constant T comparatively apace through simple iterative algorithm, be convenient on hardware view, realize.
S301: when calculating for the first time, said n is set to the initial value exponent number, promptly current exponent number n
NBe made as the initial value exponent number, according to second functional relation and current exponent number n
NCalculate the N time current inertia constant T
N
S302: confirm exponent number again according to first functional relation and the N time current inertia constant, obtain (N+1) inferior inertia exponent number n
N+1
S303: according to said second functional relation and said (N+1) inferior inertia exponent number n
N+1Again confirm inertia constant, obtain (N+1) inferior inertia constant T
N+1
S304: when the N time exponent number and said (N+1) inferior exponent number of confirming again satisfy pre-conditioned; Then (N+1) inferior exponent number is confirmed as current exponent number, and will confirm as current inertia constant according to (N+1) inferior inertia constant that (N+1) inferior exponent number obtains.
At first; First functional relation (7) is done suitable distortion; Make it be applicable to iterative algorithm; Replace
with
, replace
with
; And preferably, said n is set to initial value exponent number 2:
Above result of calculation is: the exponent number value of
inferior appearance inertia.
is the cycle calculations number of times;
is last calculation times;
is the exponent number of
inferior appearance inertia of confirming again, and
is current inertia constant.Wherein
is for calculating for the first time; Because current inertia constant is not calculated as yet, so its default value is represented with 0.
Secondly; Second functional relation (8) is done suitable distortion; Make it be applicable to iterative algorithm; Replace
with
, replace
, obtain (13) formula with
:
(13)
Above result of calculation is:
inferior appearance inertia constant calculated value.The inertia exponent number n that promptly utilizes (12) formula to obtain
NAgain confirm inertia constant T
N
Do the gain coefficient of suitable The deformation calculation system object parameter model with (9) formula; Replace
with
, replace
, replace
, obtain (14) formula with
with
:
Above result of calculation is:
inferior gain coefficient calculated value.
With (11) formula computing system image parameter model time lag constant; Replace
with
, replace
with
; Replace
with
, obtain (15) formula:
Above result of calculation is:
inferior time lag constant calculations value.
Fig. 5 is the second embodiment synoptic diagram that the transport function of system object parameter model of the present invention is confirmed method.In order specifically to illustrate the iterative algorithm of present embodiment, enumerate an instantiation below in conjunction with Fig. 5.
Supposing the system image parameter model is appearance time lag links such as one 8 rank, and the theoretical transport function of this object is:
Below the utilization computing method of the present invention to its calculate, emulation.
Select 3 identification Frequency points to be:
The amplitude transfer coefficient that obtains each identification Frequency point is:
Obtain the phase transfer value of said each identification Frequency point:
N is preset as
initial value 2 with the inertia exponent number, and n1=2 substitution (13) formula is obtained current inertia constant T1; This inertia constant T1 substitution (12) formula is confirmed exponent number again, obtain inertia exponent number n2; This inertia exponent number n2 substitution (13) formula is confirmed inertia constant again, obtain inertia constant T2; So analogize, constantly iteration., calculate inertia constant T1 according to inertia exponent number n1 for the first time and be some S1 as description of drawings with shown in Figure 5, the inertia exponent number n2 that calculates according to inertia constant T1 for the second time is some S2; The inertia constant T2 that calculates according to inertia exponent number n2 for the second time is some S3 ... So analogize; As shown in Figure 5, ordered series of numbers S1, S2; S3 ... Constantly converge on Sn, be institute and ask.Simulation calculation is as a result shown in the following table:
Can find out that from last table after cycle index was greater than 49 times, each calculation of parameter value stabilization was constant, but also have certain error that reason is to have due to the error in the identified parameters that each single-point identification frequency obtains with target transfer function theoretical parameter value.The object model transport function final calculation result that obtains is:
, the theoretical transport function of visible transport function final calculation result and system object parameter model is very approaching.
In the scope that error allows; Can obtain inertia exponent number n and inertia constant T fast through the method shown in second embodiment; When the N time inertia exponent number and said (N+1) inferior inertia exponent number of confirming again satisfy pre-conditioned; Then that (N+1) is inferior inertia exponent number is confirmed as the inertia exponent number of final value, and will confirm as the final value inertia constant according to (N+1) inferior inertia constant that (N+1) inferior inertia exponent number obtains.Preferably, for present embodiment determined (12) formula and (13) formula, saidly pre-conditionedly can comprise (14) formula:
As current inertia exponent number n
N+1Inertia exponent number n with the last time
NSatisfy (14) formula, can utilize current inertia exponent number n
N+1Calculate inertia constant T
N+1After, and stop circulation.As above shown in the table; After cycle index is greater than 8 times;
promptly satisfies condition; So, need not in the calculating of carrying out N9.
When current inertia exponent number is integer, then this inertia exponent number is confirmed as final inertia exponent number, current inertia constant is confirmed as final inertia constant;
When current inertia exponent number is not integer, to its processing that rounds up, obtain final inertia exponent number, and confirm inertia constant again according to said second functional relation and said final inertia exponent number, obtain final inertia constant;
According to the amplitude transfer coefficient of any identification Frequency point, final inertia exponent number and final inertia constant, calculate said the grade and hold the final proportional gain of link, obtain final proportional gain;
According to the phase transfer value of any identification Frequency point, final inertia exponent number and final inertia constant, calculate the time lag constant, obtain final time lag constant;
According to said final inertia exponent number, said final inertia constant, said final proportional gain, said final time lag constant, obtain the transport function of system object parameter model.
Since heat power engineering system image parameter model etc. the appearance exponent number generally can be preset as integer, so, the processing that can the final calculation result of inertia exponent number be rounded up.Confirm parameters such as final inertia constant, proportional gain, time lag constant again according to final inertia exponent number.In sum, available object model transport function final calculation result is:
.Ining contrast to original system resembles the theoretical transport function of parameter model and can know that this result of calculation is more accurate.
Need to prove;
carried out the initial value setting;
initial value scope is so long as positive integer gets final product, and reason is can finally converge on the Sn point through this iterative algorithm.Preferably,
initial value is made as 2.
In sum; The transport function of the system object parameter model that the embodiment of the invention provides is confirmed method; Have calculated amount less higher with precision, advantage such as be easy in active computer control (DCS) system, use, significant to the characteristic of accurate understanding and recognition system image parameter model.
Fig. 6 is the first embodiment synoptic diagram that the transport function of system object parameter model of the present invention is confirmed device, comprising:
The parameter model tectonic element is used for the transport function of predetermined system image parameter model, and wherein said transport function comprises that n rank etc. hold the time lag link;
Amplitude transfer coefficient acquiring unit is used to obtain the amplitude transfer coefficient of several preset identification Frequency points;
The phase transfer value acquiring unit is used to obtain the phase transfer value of several preset identification Frequency points;
With the parameter calculation unit that said amplitude transfer coefficient acquiring unit, said phase transfer value acquiring unit link to each other respectively, be used for calculating inertia exponent number, inertia constant, proportional gain, the time lag constant of said transport function according to the phase transfer value of the amplitude transfer coefficient of identification Frequency point and any identification Frequency point;
With the model acquiring unit that said parameter model tectonic element, parameter calculation unit link to each other respectively, be used for according to the get parms transport function of model of said parameter calculation unit parameters calculated.
Because thermal power plant's heat power engineering system image parameter model also has pure time-delay characteristics except having characteristics such as big inertial time sex change.So, need to confirm that system object parameter model model is appearance time lag links such as n rank, be specially:
As shown in Figure 6, also comprise first algorithm unit and second algorithm unit that link to each other with said parameter calculation unit,
Said first algorithm unit comprises: first function is confirmed the unit, is used for confirming first functional relation between inertia exponent number n and the inertia constant T; Second function is confirmed the unit, is used for confirming second functional relation between inertia exponent number n and the inertia constant T;
Said second algorithm unit comprises: the scale-up factor function is confirmed the unit, is used for confirming to calculate the functional relation of scale-up factor; The time lag constant function is confirmed the unit, is used for confirming to calculate the functional relation of time lag constant.
Fig. 7 is the second embodiment synoptic diagram that the transport function of system object parameter model of the present invention is confirmed device, and said parameter calculation unit comprises:
Initial value is provided with the unit, is used for said n is carried out the initial value setting;
With said initial value unit, said amplitude transfer coefficient acquiring unit, said second function are set and confirm the inertia constant computing unit that the unit links to each other respectively, be used for confirming inertia constant again according to the initial setting up value of said second functional relation and inertia exponent number;
Confirm the inertia exponent number computing unit that the unit links to each other respectively with said inertia constant computing unit, said amplitude transfer coefficient acquiring unit, said first function, be used for confirming exponent number again according to first functional relation and current inertia constant;
Said inertia constant computing unit also is used for confirming current inertia constant again according to said second functional relation and current inertia exponent number.
Said parameter calculation unit also comprises:
Be connected the circulation stop element between said inertia exponent number computing unit and the said inertia constant computing unit; Be used for satisfying pre-conditioned when the N time exponent number and said (N+1) inferior exponent number of confirming again; Then (N+1) inferior exponent number is confirmed as current exponent number, and will confirm as current inertia constant according to (N+1) inferior inertia constant that (N+1) inferior exponent number obtains.
Confirm the scale-up factor computing unit that unit, said final argument arrangement unit links to each other respectively with said amplitude transfer coefficient acquiring unit, said scale-up factor function, be used for the current scale-up factors of appearance link such as amplitude transfer coefficient, scale-up factor function calculation according to any identification Frequency point be said;
Confirm the time lag constant computation unit that unit, said final argument arrangement unit link to each other respectively with said phase transfer value acquiring unit, said time lag constant function, the phase transfer value, the time lag constant function that are used for according to any identification Frequency point calculate the current time lag constants of appearance link such as said;
Put the unit in order with the final argument that said inertia constant computing unit, said inertia exponent number computing unit, said circulation stop element, said scale-up factor computing unit, said time lag constant computation unit link to each other respectively; Be used for when current inertia exponent number is integer; Then this inertia exponent number is confirmed as final inertia exponent number, current inertia constant is confirmed as final inertia constant; When current inertia exponent number is not integer, to its processing that rounds up, obtain final inertia exponent number, and confirm inertia constant again according to said second functional relation and said final inertia exponent number, obtain final inertia constant;
Said scale-up factor computing unit also is used for the amplitude transfer coefficient according to any identification Frequency point, final inertia exponent number and final inertia constant, confirms the said scale-up factor that holds link that waits again, obtains to wait the final scale-up factor that holds link;
Said time lag constant computation unit also is used for the phase place transfer coefficient according to any identification Frequency point, final inertia exponent number and final inertia constant, confirms again to obtain final time lag constant by the time lag constant;
Said model acquiring unit also is used for obtaining the transport function of system object parameter model according to said final inertia exponent number, said final inertia constant, said final proportional gain, said final time lag constant.
Step in the apparatus and method of Fig. 7 is corresponding, identical in the embodiment of each unit and the method.
This paper has proposed a kind of strategy of knowing appearance Object with Time Delay Model Transfer functions such as high-order of debating based on the finite frequency performance data; And this strategy conducted in-depth analysis, provide frequency domain and debated the principle of knowledge, have the simple and stronger anti-interference of algorithm; Also has the higher knowledge precision of debating; Be easy in existing DCS, realize, to characteristics such as process influence are less, simulation study verified should strategy validity.
Through the description of above embodiment, those skilled in the art can be well understood to the present invention and can realize by the mode that software adds essential hardware platform, can certainly all implement through hardware.Based on such understanding; All or part of can the coming out that technical scheme of the present invention contributes to background technology with the embodied of software product; This computer software product can be stored in the storage medium, like ROM/RAM, magnetic disc, CD etc., comprises that some instructions are with so that a computer equipment (can be a personal computer; Server, the perhaps network equipment etc.) carry out the described method of some part of each embodiment of the present invention or embodiment.
Above-described embodiment of the present invention does not constitute the qualification to protection domain of the present invention.Any modification of within spirit of the present invention and principle, being done, be equal to replacement and improvement etc., all should be included within the claim protection domain of the present invention.