CN101458521B - Misoperation prevention method by DCS - Google Patents

Abstract

The invention relates to a method of DCS preventing misoperation which includes steps as follows: firstly, determining a computing formula to boundary value SHk, SHk and limiting value SH, SL by using regression computing method and temporal analysis: recording initialize upper limit and down limit in DCS parameter as an initialize value before putting limit switch; secondly, judging amend given VS value whether in limiting range; thirdly, judging counter whether equal to pre-set counting value, namely judging whether delay arrival time; fourthly, judging boundary value SHk and SLk and amending SH and SL. The method can hold up most of misoperation by operator by using advanced control, intelligent operation judgement and misoperation holding up alarm prompt system, ensures production safety, and can enhances enterprise's economic efficiency by using long period production. The method provided by the invention can be used in various petrochemical operation produce.

Description

A kind of DCS prevents the method for maloperation

Technical field

The present invention relates to a kind of DCS (Distributed Control System (DCS)) method of operating, particularly about a kind of method that prevents the DCS maloperation of set-point (SV value) in proportion integration differentiation control (PID) loop of being used for of in advanced person control (APC) intellectualized technology field, revising.

Background technology

In Chemical Manufacture, widely-used DCS carries out the automatic control of production technology.In the DCS system, when the operative employee controls the SV value modification in (PID) loop to a proportion integration differentiation, because the size of SV value has determined the pid loop output controlling value size of PV value (PV is the pid loop measured value) amplitude amplitude just, do not exceed when the SV value under the situation of upper limit SH, lower limit SL scope, the DCS system can not do any judgement and carry out; When the SV value exceeds upper limit SH or lower limit SL, system will eject the alarm dialog box, and operation can be selected to continue to carry out or interrupt carrying out, and upper limit SH or lower limit SL are traditionally arranged to be technological limits parameter (as shown in Figure 6).Prior art is that the control method that high and low limit (being upper and lower limit) is carried out the edge amplitude limit is exported in the employing that skies company uses, so that do not exceed edge amplitude limit scope during operation, to protect handling safety.But in the high and low limited field of output, once revise given amplitude then without limits, so under many circumstances, because the operative employee's is careless, once revises and give the fixed step size amplitude excessive, under the effect of PID, can bring bigger overshoot to system, thereby security of system is brought very big danger, gently then cause system shutdown, heavy then can cause security incident.

Summary of the invention

At the problems referred to above, the purpose of this invention is to provide and a kind ofly can carry out intelligent decision given SV value step-length amplitude once, make the reliable DCS of security of system prevent the method for maloperation.

For achieving the above object, the present invention takes following technical scheme: a kind of DCS prevents the method for maloperation, and it may further comprise the steps: step 1, utilize regressing calculation method and temporal analysis to determine boundary value SH _k, SL _kComputing formula with amplitude limit value SH, SL: before dropping into the amplitude limit switch, initial setting upper limit magnitude and the Lower Limit Amplitude write down in the DCS parameter are initial set value, and according to technology with reference to the calculation of parameter value of revising the boundary: the boundary value SH of upper bound scope _k=SH '-K _pA; The boundary value SL of lower bound scope _k=SL '+K _pA, wherein A be technology with reference to parameter, K _pIt is correction factor; At the amplitude limit switch is amplitude limit switch when dropping into for the first time: revising given maximum limit amplitude SH is SV+K/b ₁-a ₁Revising given irreducible minimum amplitude SL is SV-K/b ₁-a ₁ Step 2, judge to revise given SV value whether in limited range: when the described amplitude limit switch of input, if the given amplitude of given described SV value goes up the amplitude range value or during less than described amplitude range value down greater than described, the DCS system sends warning automatically and does not carry out, described SV value is that given preceding initial value is constant, original state is returned in the time-delay of not enabling counting device; If the given amplitude of given described SV value is less than described upward amplitude range value or greater than described amplitude range value down, in the time of in limited amplitude range, described SV value is this given SV value, start described counter time-delay; After described amplitude limit switch disconnects, then described upper and lower scope amplitude limit value is returned to described initial value; Step 3, judge that whether counter equals to set count value, promptly judges whether delay arrival time: when described Counter Value equals to set count value, just revise delay time then, enter boundary value and judge; When described Counter Value was not equal to the setting count value, just described correction time-delay was not then then returned original state, re-executes order, continues to wait for, equals to set count value up to described Counter Value; Step 4, judgement boundary value SH _kAnd SL _kAnd revise SH and SL: when described Counter Value equals to set count value, just revise delay time then, judge that whether upper boundary values is greater than SV+K/b earlier ₁-a ₁If described boundary value is greater than SV+K/b ₁-a ₁, then revise given amplitude maximal value and equal SV+K/b ₁-a ₁, enter lower border value and judge; If described upper boundary values is less than SV+K/b ₁-a ₁, the given amplitude maximal value of then described correction equals described upper boundary values, enters described lower border value and judges; If described lower border value is less than SV-K/b ₁-a ₁, then revise given amplitude minimum value and equal SV-K/b ₁-a ₁, described counter O reset; If described lower border value is greater than SV-K/b ₁-a ₁, the given amplitude minimum value of then described correction equals described lower border value, and original state is returned in described counter O reset.

Described amplitude limit switch is a soft switch that utilizes software to realize.

Described temporal analysis comprises first-order system and second-order system.

The described step size increments and the described maximum overshoot mathematical equation that utilize regressing calculation to obtain are: h _Max=b ₁Δ SV+a ₁

H wherein _MaxMaximum overshoot, Δ SV are step size increments, b ₁And a ₁Be constant; Utilize regressing calculation to obtain described step size increments and the mathematical equation of described adjusting time is:

T _S＝b ₂·ΔSV+a ₂

T wherein _SBe adjusting time, b ₂And a ₂Be constant.

Under the prerequisite of described maximum overshoot smaller or equal to the maximum variable quantity of circuit controls value PV of described pid loop, the given described amplitude maximal value or the computing method of given described amplitude minimum value and the time interval and boundary value are: given amplitude maximal value SH=SV+K/b ₁-a ₁Given amplitude minimum value SL=SV-K/b ₁-a ₁Revise the time interval T of step size increments Δ SV _Sp=b ₂(K/b ₁-a ₁)+a ₂The boundary value SH of upper bound scope _k=SH '-K _pA; The boundary value SL of lower bound scope _k=SL '+K _pA, wherein A be technology with reference to parameter, K _pIt is correction factor.

The present invention is owing to take above technical scheme, it has the following advantages: 1, the present invention is owing to carry out intelligent decision the operative employee once being revised given SV value step-length amplitude, therefore super amplitude range then alarm point out and do not carry out, not super amplitude range is then only carried out in SL～SH amplitude range, and reaches upper boundary values SH when the SV value _kOr lower border value SL _kDuring scope, then can not operate execution up and down automatically.2, the present invention is equivalent to disturbance of pid loop owing to the given SV value of step, therefore the given SV value of step step-length amplitude is big more as can be known, then amplitude is big more in the pid loop amplitude process, and in the temporal analysis of the calculating employing automatic control system of carrying out control SL～SH amplitude range, the adjusting time T of the dynamic perfromance of data acquistion system by experiment _sWith maximum overshoot h _Max, determine step given step size increments Δ SV and maximum overshoot h according to the one-variable linear regression method _MaxAnd adjusting time T _sFuntcional relationship, thereby calculate SL and the upper and lower bounds value of SH SH _kAnd SL _k, so SH _k, SL _kValue can be done according to process system controlled variable A and on-the-fly modify.3, the present invention is because DCS adopts an amplitude limit switch (being soft switch) control operation, make fool proof input and withdraw from and obtained effective control, and the amplitude limit switch writes down former SL, SH automatically when not dropping into initial value is SH ' and SL ', drops into and withdraws from the value that original SL, SH are recovered in the back.4, the present invention is owing to after the change of SV Value Operations, postpone to regulate time T _sAfter, can be once more to the operation of making amendment of SV value, and revise the value of SL, SH automatically, make the upper and lower limit amplitude range can follow the tracks of variation.In sum, the present invention has realized that intellectualized operation is judged and maloperation interception alarm, therefore most of maloperation of operative employee is blocked, and has ensured the security of producing, and can satisfy the requirement that long period is produced, and can improve economic benefit of enterprises.

The present invention can be widely used in the DCS application system of various petrochemical compleies productions.

Description of drawings

Fig. 1 is the limited synoptic diagram of opereating specification of the present invention

Fig. 2 is a first-order system mathematical model synoptic diagram of the present invention

Fig. 3 is the unit-step response curve synoptic diagram of first-order system of the present invention

Fig. 4 is a second-order system mathematical model synoptic diagram of the present invention

Fig. 5 is the unit-step response curve synoptic diagram of second-order system of the present invention

Fig. 6 is a prior art system opereating specification synoptic diagram

Embodiment

Below in conjunction with drawings and Examples the present invention is described in detail.

In the pid loop of DCS (Distributed Control System (DCS)), because when the given SV value of step function, loop output control PV value (PV is the loop measured value) an amplitude process can occur under the effect of pid parameter, be equivalent to the disturbance in loop, when given SV value step-length amplitude is big more, the amplitude of PV value is just big more, so the operative employee controls the amplitude of PV value by once revising given SV value amplitude.In order to improve the reliability of DCS, avoid maloperation, the present invention prevents the DCS maloperation by the amplitude of once revising given SV value is carried out intelligent decision.

As shown in Figure 1, the limited operation scope of given SV value comprises the boundary value SH of the upper and lower boundary scope of solid line part _k, SL _k, boundary value SH _k, SL _kCan do according to the process system controlled variable and on-the-fly modify, for example technology is with reference to the parameter A value.Apart from boundary value SH _k, SL _kUpper and lower edge dotted portion be respectively the initial value SH ' of system and the SL ' of upper and lower amplitude range value, apart from SH _kAnd SL _kBetween upper and lower ladder solid line be respectively amplitude range value SH and following amplitude range value SL, the ladder dotted line between amplitude range value SH and SL is pid loop modification set-point SV.What the present invention will realize is, in the DCS system, and when given SV value surpasses amplitude range SL～SH, auto-alarm-signal keying device prompt facility then.Specifically be divided into, when the SV value surpasses amplitude range SL～SH, then alarm prompting and not carrying out; When the SV value is no more than amplitude range SL～SH, then only in amplitude range SL～SH, carry out; When given SV value reaches upper boundary values SH _kOr lower border value SL _kDuring scope, then can not operate execution up or down automatically.

Because the amplitude of PV value is big or small relevant with the dynamic perfromance of SV value and pid loop, the state that finally can tend towards stability of the amplitude process in pid loop, hence one can see that, and the dynamic analysis of pid loop is relevant with the dynamic analysis of automatic control system.The dynamic analysis of automatic control system has temporal analysis, root-locus technique and frequency domain analysis, these three kinds of methods commonly used are analyzed the performance of linear control system, but diverse ways has the different characteristics and the scope of application, comparatively speaking, temporal analysis is a kind of method of directly in time domain system being analyzed, has intuitive and accurate advantage, and can provide the full detail of time response, be applicable to experimental technique, very easily the characterisitic parameter of controlled system.Therefore the present invention adopts the temporal analysis of automatic control system, the dynamic Characteristic Data of data acquistion system by experiment, and step response functions is the PV value, even h (t)=PV.Specify the analytic process of temporal analysis below:

One, typical original state analysis: the original state of stipulating system is a zero condition, that is:

$c\left(0^{-}\right)=\stackrel{\·}{c}\left(0^{-}\right)=\stackrel{\·\·}{c}\left(0^{-}\right)=\·\·\·=0.$

Two, typical input signal: typical input signal is actual input approximate and abstract of numerous complicated, is meant the input signal form that often runs into according to system, more Utopian basic input functions in addition on mathematical description.The selection of typical case's input signal should make mathematical operation simple, is convenient to experimental verification again, and typical input signal commonly used has unit-step function:

$1\left(t\right)=\left\{\begin{array}{cc}1& t>0\\ 0& t\≤0\end{array}\right.$

Three, typical time response: the typical case is that original state is zero the output of system under typical input signal effect time response, it is made up of dynamic process and steady-state process two parts: dynamic process claims transient process or transient process again, the system that is meant is under typical input signal effect, and system's output is arrived the response process of end-state by original state; Steady-state process is meant system under the effect of typical input signal, when the time, t was tending towards infinity, and the form of expression of system's output quantity.

Four, the analysis of dynamic property and steady-state behaviour: the performance index of control system under typical input signal effect, form by dynamic property and steady-state behaviour two parts.

Dynamic property is meant describes stable system under the unit-step function effect, and dynamic process is the index of the changing condition of t in time, and the dynamic performance index of unit-step response has usually: t during delay _d, refer to that response curve reaches half needed time of its final value for the first time; Rise time t _r, refer to that response curve rises to the 90% needed time of final value from final value 10%, for the system that vibration is arranged, also may be defined as response and rise to the needed time of final value from zero first time, the rise time is a kind of tolerance of system response time; Time to peak t _p, refer to that response surpasses final value and reaches the needed time of first peak value; Adjusting time t _s, refer to that response reaches and remains on the needed time in final value ± 5% (or ± 2%); Maximum overshoot h _Max, refer to the maximum deviation amount h (t that responds _p) poor with final value h (∞).

Steady-state error is a kind of performance index of descriptive system steady-state behaviour.Measure or calculate under step function, ramp function and acceleration function effect, steady-state error is a kind of tolerance of system's control accuracy or disturbance rejection ability.When the time was tending towards infinity, the output quantity of system was not equal to definite function of input quantity or input quantity, and then there is steady-state error in system.

The present invention adopts first-order system and second-order system as input signal in the pid loop of DCS, and h (t)=PV can release the time-domain analysis of first-order system as follows by above-mentioned temporal analysis:

The mathematical model of first-order system: as shown in Figure 2, the open-loop transfer function of system is $G\left(s\right)=\frac{1}{\mathrm{Ts}}=\frac{K}{s},$ The closed loop transfer function, of system is $\mathrm{\Φ}\left(s\right)=\frac{K}{s+K}=\frac{1}{\mathrm{Ts}+1}.$

The response of first-order system: the present invention utilizes unit-step response, when input signal is unit step signal, i.e. and r (t)=1 (t), its Lars is transformed to $R\left(s\right)=\frac{1}{s},$ Then have:

$C\left(s\right)=\mathrm{\Φ}\left(s\right)R\left(s\right)=\frac{1}{s(\mathrm{Ts}+1)}=\frac{1}{s}-\frac{1}{s+1/T}$

As shown in Figure 3, following formula is got the Lars inversion, can obtain the unit-step response of first-order system, its function is:

h(t)＝L ^-1[C(s)]＝1-e ^-t/T

According to the definition of dynamic performance index, can obtain dynamic performance index t time delay of first-order system _d=0.69T, rise time t _r=2.20T regulates time t _s=3T (± 5%).For the unit-step response of first-order system, steady-state error $e_{\mathrm{ss}}=\underset{t\→\∞}{\lim }e\left(t\right)=\underset{t\→\∞}{\lim }[r\left(t\right)-c\left(t\right)]=0,$ When being first-order system tracking step input signal, no steady-state error.

When in pid loop, adopting second-order system, can release time-domain analysis to second-order system by above-mentioned temporal analysis as input signal.

The mathematical model of typical case's second-order system: as shown in Figure 4, the open-loop transfer function of system is $G\left(s\right)=\frac{{\mathrm{\&omega;}}_n^2}{s(s+2\mathrm{\&xi;}{\mathrm{\&omega;}}_n)}=\frac{K}{s(\mathrm{Ts}+1)},$ The closed loop transfer function, of system is: $\mathrm{\&Phi;}\left(s\right)=\frac{{{\mathrm{\&omega;}}_n}^2}{{\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="H2008102473682E00021.png"\; state="\{\{state\}\}">s</figure-callout>}}^2+2\mathrm{\&xi;}{\mathrm{\&omega;}}_{n}s+{{\mathrm{\&omega;}}_{\mathrm{<figure-callout\; id="2"\; label="n"\; filenames="H2008102473682E00021.png"\; state="\{\{state\}\}">n</figure-callout>}}}^2},$ Wherein ξ is a damping ratio.

The unit-step response of second-order system: second order unit-step response function h (t) comprises steady-state component and dynamic component, and wherein steady-state component is 1, and dynamic component presents oscillatory extinction characteristic (as shown in Figure 5).The steady-state error e of second-order system _Ss=0, when promptly typical second-order system was followed the tracks of step input signal, no steady-state error, system were astatic system.According to the definition of dynamic performance index, can obtain following two conclusions:

Conclusion one: damping ratio ξ is big more, and the overshoot of system is more little, and response steadily; Damping ratio ξ is more little, and the overshoot of system is big more, and the stationarity of response is poor more; When ξ=0, the response of system is h (t)=1-cos ω _nT, h (t) is ω for frequency _nContinuous oscillation, system can't enter balance duty, cisco unity malfunction; As ξ one timing, ω _nBig more, the oscillation frequency ω of system _dBig more, the stationarity of response is relatively poor.Therefore it is big to work as damping ratio ξ, frequencies omega _nHour, the stationarity of system responses is good.

Conclusion two: regulate time t _sComputing formula be approximate expression, in fact, when damping ratio ξ hour, the system responses speed of convergence was slow, regulated time t _sLong; When damping ratio ξ was excessive, system responses was blunt, regulated time t _sAlso longer.Therefore damping ratio ξ should get suitable numerical value, and the typical second-order system when damping ratio ξ=0.707 is called best second-order system, this moment overshoot h _MaxBe 4.3%, regulate time t _sBe 3/ ω _n

Because given SV value is given by step function, in the pid loop of DCS, adopt first-order system and second-order system as input signal, can be similar to the step given step size increments Δ SV and the maximum overshoot h that think near given SV value _MaxAnd adjusting time T _SThere is linear relationship.Can determine step given step size increments Δ SV and maximum overshoot h by the one-variable linear regression method _MaxAnd adjusting time T _SFuntcional relationship, and determine computing method and the time interval T of given amplitude maximal value SH or given amplitude minimum value SL _SpWith boundary value SH _k, SL _kExperimental method realizes that obtaining the regressing calculation data is in output loop, constantly to the given different range value of SV value, measures each maximum overshoot h _MaxWith the adjusting time T _S, draw some groups of data and carry out regressing calculation; In actual production process, realize obtaining the regressing calculation data and be data, analytic record is carried out in operative employee's the given operation of SV value, draw some groups of data and carry out regressing calculation by host computer opc server interface interchange DCS.Utilize the regressing calculation method can on-line measurement, and can remeasure calculating, make amendment calculating regression formula every a period of time.Utilize regressing calculation to calculate the given step size increments Δ SV that jumps and specifically comprise following step:

1) can determine to regulate time T according to the one-variable linear regression equation _SWith the pass of step input signal step size increments Δ SV be:

If T _S=f (Δ SV) regulates time T _SFunction be the function of step input signal step size increments Δ SV, some groups of n some experimental data (Δ SV are arranged ₁, T _S1), (Δ SV ₂, T _S2) ..., (Δ SV _n, T _Sn).

Order ${\hat{T}}_S=a+\mathrm{b\&Delta;SV}$ Be formula (1) that wherein a is a straight line intercept constant, b is the straight slope constant, regulates the estimated value of time

With step size increments Δ SV be linear relationship, can determine the value of constant a and b according to least square method:

Order $Q(a,b)=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{[T_{\mathrm{Si}}-(a+b{\left(\mathrm{\&Delta;SV}\right)}_i)]}^2,$ Then (a b) should satisfy at the limit place Q $\left\{\begin{array}{c}\frac{\&PartialD;Q}{\&PartialD;a}=0\\ \frac{\&PartialD;Q}{\&PartialD;b}=0\end{array}\right.,$ That is:

$\left\{\begin{array}{c}\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}[T_{\mathrm{Si}}-(a+b{\left(\mathrm{\&Delta;SV}\right)}_i)]=0\\ \underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}[T_{\mathrm{Si}}-(a+b{\left(\mathrm{\&Delta;SV}\right)}_i)]{\left(\mathrm{\&Delta;SV}\right)}_i=0\end{array}\right.$

Can get the system of linear equations of a, b by following formula, be designated as formula (2):

$\left\{\begin{array}{c}\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{T}_{\mathrm{Si}}-\mathrm{na}-b\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\mathrm{\&Delta;SV}\right)}_i=0\\ \underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\mathrm{\&Delta;SV}\right)}_i{T}_{\mathrm{Si}}-a\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\mathrm{\&Delta;SV}\right)}_i-b\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\mathrm{\&Delta;SV}\right)}_i^2=0\end{array}\right.$

By above-mentioned normal equations can declare know Q (a b) has minimum point, separates normal equations and can obtain a and b one group and separate, be designated as (

), can demonstrate,prove (

) be Q (a, minimum point b), then

With Formula be respectively formula (3) and formula (4).

Formula (3) is: $\hat{a}=\stackrel{\&OverBar;}{T_S}-\hat{b}\stackrel{\&OverBar;}{\mathrm{\&Delta;SV}}.$

Formula (4) is: $\hat{b}=\frac{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\mathrm{\&Delta;SV}\right)}_i{T}_{\mathrm{Si}}-n\stackrel{\&OverBar;}{{\mathrm{\&Delta;SVT}}_S}}{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\mathrm{\&Delta;SV}\right)}_i^2-n\stackrel{\&OverBar;}{{\left(\mathrm{\&Delta;SV}\right)}^2}}=\frac{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}({\left(\mathrm{\&Delta;SV}\right)}_i-\stackrel{\&OverBar;}{\mathrm{\&Delta;SV}})(T_{\mathrm{Si}}-\stackrel{\&OverBar;}{T_S})}{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{({\left(\mathrm{\&Delta;SV}\right)}_i-\stackrel{\&OverBar;}{\mathrm{\&Delta;SV}})}^2}.$

In the following formula $\stackrel{\&OverBar;}{\mathrm{\&Delta;SV}}=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\mathrm{\&Delta;SV}\right)}_i,$ $\stackrel{\&OverBar;}{T_S}=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{T}_{\mathrm{Si}}.$

Formula (3) and (4) formula are determined

Can obtain formula (5) in the substitution formula (1) is:

${\hat{T}}_S=\hat{a}+\hat{b}\mathrm{\&Delta;SV}$

Formula (5) is exactly step size increments Δ SV and adjusting time T _SBetween the correlation experience formula, be called regression beeline equation.

2) to regulating time T _SCarry out correlation test, have:

Residual sum of squares (RSS): $Q=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(T_{\mathrm{Si}}-{\hat{T}}_{\mathrm{Si}})}^2;$

Regression sum of square: $U=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{({\hat{T}}_{\mathrm{Si}}-\stackrel{\&OverBar;}{T_S})}^2;$

Related coefficient: $R=\frac{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{({\hat{T}}_{\mathrm{Si}}-\stackrel{\&OverBar;}{T_s})}^2}{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(T_{\mathrm{Si}}-\stackrel{\&OverBar;}{T_S})}^2};$

In the above-mentioned formula, when residual sum of squares (RSS) Q was big, then meter sensitivity increased, otherwise meter sensitivity diminishes.When coefficient R is big, then there is linear dependence; When R hour, then do not have linear dependence.

3) can determine maximum overshoot h according to the one-variable linear regression equation _MaxWith the pass of step input signal step size increments Δ SV be:

If h _Max=f (Δ SV), i.e. maximum overshoot h _MaxFunction be the function of step input signal step size increments Δ SV, have n the point some groups of experimental data (Δ SV ₁, h _Max1), (Δ SV ₂, h _Max2) ..., (Δ SV _n, h _Maxn).

Order ${\hat{h}}_{\max }=a+\mathrm{b\&Delta;SV}$ Be formula (6) that wherein a is a straight line intercept constant, b is the straight slope constant, the estimated value of maximum overshoot

With step size increments Δ SV be linear relationship, can determine the value of constant a and b according to least square method, solution procedure and above-mentioned definite adjusting time T _SIdentical with the value method of finding the solution constant a and b in the step input signal increment Delta SV relation, therefore can obtain formula (7) and be:

${\hat{h}}_{\max }=\hat{a}+\hat{b}\mathrm{\&Delta;SV}$

Formula (7) is exactly step size increments Δ SV and maximum overshoot h _MaxBetween the correlation experience formula, be called regression beeline equation.

4) to maximum overshoot h _MaxCarry out correlation test, have:

Residual sum of squares (RSS): $Q=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(h_{\max i}-{\hat{h}}_{\max i})}^2$

Regression sum of square: $U=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{({\hat{h}}_{\max i}-\stackrel{\&OverBar;}{h_{\max }})}^2$

Related coefficient: $R=\frac{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{({\hat{h}}_{\max i}-\stackrel{\&OverBar;}{h_{\max }})}^2}{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(h_{\max i}-\stackrel{\&OverBar;}{h_{\max i}})}^2}$

In the above-mentioned formula, when residual sum of squares (RSS) Q was big, then meter sensitivity increased, otherwise meter sensitivity diminishes.When coefficient R is big, then there is linear dependence; When R hour, then do not have linear dependence.

By above-mentioned 1)～4) the step analytical calculation, according to experimental data,, can obtain maximum overshoot h by regression beeline equation (7) _MaxMathematical equation with step size increments Δ SV is designated as: h _Max=b ₁Δ SV+a ₁By regression beeline equation (5), can adjusted time T _SMathematical equation with step size increments Δ SV is designated as: T _S=b ₂Δ SV+a ₂According to the requirement of process system, the maximum overshoot h of pid loop _MaxShould≤K, K is the maximum variable quantity of the circuit controls value PV of technological requirement, and K value determined by the technology personnel, so can determine computing method and the time interval T of given amplitude limit maximal value SH or given amplitude limit minimum value SL _SpWith boundary value SH _k, SL _k, then:

Given amplitude limit maximal value SH=SV+K/b ₁-a ₁

Given amplitude limit minimum value SL=SV-K/b ₁-a ₁

Revise the time interval T of step size increments Δ SV _Sp=b ₂(K/b ₁-a ₁)+a ₂

The boundary value SH of upper bound scope _k=SH '-K _pA;

The boundary value SL of lower bound scope _k=SL '+K _pA,

Wherein A be technology with reference to parameter, K _pBe correction factor, K _pCan set according to actual conditions and experience.Because first-order system does not have overshoot h _Max, can set the scope of initial set value SH and SL according to actual conditions, regulate time T _SCan draw through the reality test according to actual conditions.

Comprehensive above explanation can obtain the following concrete operations step that amplitude that the present invention once revises given SV value is carried out intelligent decision:

Step 1, utilize the temporal analysis of regressing calculation method and automatic control system to determine to boundary value SH _k, SL _kComputing formula with amplitude limit value SH, SL: amplitude limit switch (for a soft switch that utilizes software to realize) is to drop into and the operation of withdrawing from anti-misoperation software in order to control.Before dropping into the amplitude limit switch, the initial setting upper limit magnitude SH, the Lower Limit Amplitude SL that write down in the DCS parameter are SH ' and SL ', and according to technology with reference to the calculation of parameter value of revising the boundary SH _kAnd SL _kBe respectively:

The boundary value SH of upper bound scope _k=SH '-K _pA;

The boundary value SL of lower bound scope _k=SL '+K _pA,

Wherein A be technology with reference to parameter, K _pBe correction factor, and K _pCan set according to actual conditions and experience.

At the amplitude limit switch is amplitude limit switch when dropping into for the first time:

Revising given maximum limit amplitude SH is SV+K/b ₁-a ₁

Revising given irreducible minimum amplitude SL is SV-K/b ₁-a ₁

Step 2, judge to revise set-point (SV value) whether in limited range: when input amplitude limit switch is the amplitude limit switch, if the given amplitude of given SV value is during greater than last amplitude range value SH or less than following amplitude range value SL, the DCS system sends warning automatically and does not carry out, the SV value is that given preceding initial value is constant, original state is returned in the time-delay of not enabling counting device; If the given amplitude of given SV value is less than last amplitude range value SH or greater than following amplitude range value SL, in the time of in limited amplitude range, the SV value is this given SV value, the time-delay of enabling counting device.

When the amplitude limit switch is after the amplitude limit switch disconnects, then upper and lower scope amplitude limit value SH and SL to be returned to initial value SH ' and SL '.

Step 3, judge that whether counter equals to set count value, promptly judges whether delay arrival time: when Counter Value equals to set count value, just revise delay time then, enter boundary value and judge; When Counter Value is not equal to the setting count value, just revise time-delay not then, then return original state, re-execute order, continue to wait for, equal to set count value up to Counter Value.

Step 4, judgement boundary value SH _kAnd SL _kAnd revise SH and SL: when Counter Value equals to set count value, just revise delay time then, judge upper boundary values SH earlier _kWhether greater than SV+K/b ₁-a ₁If, upper boundary values SH _kGreater than SV+K/b ₁-a ₁, then revise given amplitude maximal value SH=SV+K/b ₁-a ₁, enter lower border value SL _kJudge; If upper boundary values SH _kLess than SV+K/b ₁-a ₁, then revise given amplitude maximal value SH=SH _k, enter lower border value SL _kJudge.If lower border value SL _kLess than SV-K/b ₁-a ₁, then revise given amplitude minimum value SL=SV-K/b ₁-a ₁, described counter O reset; If lower border value SL _kGreater than SV-K/b ₁-a ₁, then revise given amplitude minimum value SL=SL _k, original state is returned in described counter O reset.

By the various embodiments described above as can be seen, the present invention comes control operation by adopting an amplitude limit switch, automatically writing down former upper and lower amplitude range value SH and SL when the amplitude limit switch does not drop into is initial set value SH ' and SL ', amplitude limit switch input is withdrawn from the back and is recovered original amplitude range value SL, SH, the input of amplitude limit switch of the present invention and the setting of withdrawing from effective control data.And owing to after the change of SV Value Operations, postpone to regulate time T _sAfter, can revise the value of SL, SH automatically, and the amplitude limit scope can be followed the tracks of variation.

Claims (6)

1. a DCS prevents the method for maloperation, and it may further comprise the steps:

Step 1, utilize regressing calculation method and temporal analysis to determine to upper boundary values SH _k, lower border value SL _kComputing formula with upper limit magnitude SH, Lower Limit Amplitude SL: before dropping into the amplitude limit switch, the initial setting upper limit magnitude SH and the Lower Limit Amplitude SL that write down in the DCS parameter are initial set value, and according to technology with reference to the calculation of parameter value of revising the boundary:

Upper boundary values SH _k=SH '-K _pA;

Lower border value SL _k=SL '+K _pA,

Wherein A be technology with reference to parameter, K _pBe correction factor, SH ' is system's initial value of upper limit magnitude SH, and SL ' is system's initial value of Lower Limit Amplitude SL;

At the amplitude limit switch is amplitude limit switch when dropping into for the first time:

The maximum limit amplitude of revising given upper limit magnitude SH is SV+K/b ₁-a ₁

The irreducible minimum amplitude of revising given Lower Limit Amplitude SL is SV-K/b ₁-a ₁

Wherein SV is that step is given, and K is the maximum variable quantity of the circuit controls value PV of technological requirement, b ₁, a ₁Be constant;

Step 2, judge to revise the given SV value of step whether in limited range: when the described amplitude limit switch of input, if the given amplitude of the given SV value of given described step is during greater than the maximum limit amplitude of described given upper limit magnitude SH or less than the irreducible minimum amplitude of described given Lower Limit Amplitude SL, the DCS system sends warning automatically and does not carry out, the given SV value of described step is that given preceding initial value is constant, original state is returned in the time-delay of not enabling counting device; If the given amplitude of the given SV value of given described step is less than the maximum limit amplitude of described given upper limit magnitude SH or greater than the irreducible minimum amplitude of described given Lower Limit Amplitude SL, in the time of in limited amplitude range, the given SV value of described step is the given SV value of this given step, starts described counter time-delay;

After described amplitude limit switch disconnects, then the maximum limit amplitude of described given upper limit magnitude SH, the irreducible minimum amplitude of given Lower Limit Amplitude SL are returned to described initial value;

Step 3, judge that whether counter equals to set count value, promptly judges whether delay arrival time: when described Counter Value equals to set count value, just revise delay time then, enter boundary value and judge; When described Counter Value was not equal to the setting count value, just described correction time-delay was not then then returned original state, re-executes order, continues to wait for, equals to set count value up to described Counter Value;

Step 4, judgement upper boundary values SH _kWith lower border value SL _kAnd revise upper limit magnitude SH and Lower Limit Amplitude SL: when described Counter Value equals to set count value, just revise delay time then, judge upper boundary values SH earlier _kWhether greater than SV+K/b ₁-a ₁If, described boundary value SH _kGreater than SV+K/b ₁-a ₁, the maximum limit amplitude of then revising given upper limit magnitude SH equals SV+K/b ₁-a ₁, enter lower border value SL _kJudge; If described upper boundary values SH _kLess than SV+K/b ₁-a ₁, the maximum limit amplitude of the given upper limit magnitude SH of then described correction equals described upper boundary values SH _k, enter described lower border value SL _kJudge; If described lower border value SL _kLess than SV-K/b ₁-a ₁, the irreducible minimum amplitude of then revising given Lower Limit Amplitude SL equals SV-K/b ₁-a ₁, described counter O reset; If described lower border value SL _kGreater than SV-K/b ₁-a ₁, the irreducible minimum amplitude of the given Lower Limit Amplitude SL of then described correction equals described lower border value SL _k, original state is returned in described counter O reset.

2. a kind of DCS as claimed in claim 1 prevents the method for maloperation, it is characterized in that: described amplitude limit switch is a soft switch that utilizes software to realize.

3. a kind of DCS as claimed in claim 1 prevents the method for maloperation, it is characterized in that: described temporal analysis comprises first-order system and second-order system.

4. a kind of DCS as claimed in claim 2 prevents the method for maloperation, it is characterized in that: described temporal analysis comprises first-order system and second-order system.

5. prevent the method for maloperation as claim 1 or 2 or 3 or 4 described a kind of DCS, it is characterized in that: the step size increments and the maximum overshoot mathematical equation that utilize regressing calculation to obtain are:

h _max＝b ₁·ΔSV+a ₁

H wherein _MaxMaximum overshoot, Δ SV are step size increments, b ₁And a ₁Be constant; Utilize regressing calculation to obtain step size increments to be with the mathematical equation of regulating the time:

T _S＝b ₂·ΔSV+a ₂

T wherein _SBe adjusting time, b ₂And a ₂Be constant.

6. a kind of DCS as claimed in claim 5 prevents the method for maloperation, it is characterized in that: under the prerequisite of described maximum overshoot smaller or equal to the maximum variable quantity of circuit controls value PV of pid loop, the computing method of the maximum limit amplitude of given described given upper limit magnitude SH, the irreducible minimum amplitude of given Lower Limit Amplitude SL and the time interval and boundary value are:

Maximum limit amplitude=SV+K/b of given upper limit magnitude SH ₁-a ₁

Irreducible minimum amplitude=SV-K/b of given Lower Limit Amplitude SL ₁-a ₁

Revise the time interval T of step size increments Δ SV _Sp=b ₂(K/b ₁-a ₁)+a ₂

The boundary value SH of upper bound scope _k=SH '-K _pA;

The boundary value SL of lower bound scope _k=SL '+K _pA,

Wherein A be technology with reference to parameter, K _pIt is correction factor.

Images