CN103760770B - Distribution type generalized forecast control method based on positive and negative input system - Google Patents

Abstract

The invention discloses distribution type generalized forecast control method based on positive and negative input system, include following steps: obtain the positive and negative input of this system and positive and negative input parameter；It is single input parameter by positive and negative input parameter synthesis；Utilize GPC algorithm and single input parameter to align, reverse pumping enters to process, thus obtains single input；Time lag according to the equivalence relation between positive and negative input and single input and positive and negative input calculates the expection positive input of positive and negative input system or expection reverse pumping enters, and then aligns anti-input system and be controlled.The present invention is by being single input parameter by positive and negative input parameter synthesis, the time lag characteristic utilizing GPC algorithm and positive and negative input calculates, control process is reflected in single input parameter, thus realize pinpoint accuracy, and reduce the energy during controlling or the consumption of material.The present invention can be widely applied to Industry Control as a kind of distribution type generalized forecast control method based on positive and negative input system.

Description

Distribution type generalized forecast control method based on positive and negative input system

Technical field

The present invention relates to industrial control method, be based especially on the distribution type generalized predictive control side of positive and negative input system Method.

Background technology

Positive and negative input system is widely present in Industry Control, refers in a control system, have two run counter to defeated Entering signal, positive input system makes output produce positive-effect, and negative input system makes output produce negative effect.The existence of this system is Because controller needs positive input to obtain the result wanted sometimes, but sometimes needs negative input unnecessary to remove Energy.

Major part technology, is the most also limited to separately consider, even if entirety is put two positive and negative input variables at present Consider also not account for the characteristic of different input variables together.For the control of positive and negative input system, mainly there are two classes, one Class is individually to control for positive and negative two different controllers of two In-put designs, typically has: double PID control, for two not Same control system, separately designs a set of PID controller and carries out control respectively.Double PID are controlled, in certain condition Under, reasonable control result can be obtained.But due to positive and negative input in this case, can open simultaneously, thus shine into energy The waste in source.This is primarily due to these methods does not has positive and negative two inputs to integrate consideration.

An other class is Staged cotrol for the method for positive and negative input system, such as PID type Staged cotrol.This kind of method sets The controller of meter is summarized as one positive and negative input and handles variable, and then design PID controller controls, at output result Which interval to distribute control positive input in or reverse pumping enters.For PID type Staged cotrol, this control method is due to time each Carve only one of which input can be opened, the energy can be saved, but owing to not accounting for positive and negative input to the response characteristic of output be Different, particularly manipulation variable switches near zero when, the control result obtained can deteriorate.So this is a kind of By the method that infringement control performance is energy-conservation.In a lot of production processes, low precision controlling is the quality that can damage product 's.Such as: in plastic extrusion production process, first barrel is preheated, when temperature arrives setting value, electricity can be started Machine, at this moment raw material can enter barrel along hopper, and under the shear action of screw rod and barrel, material melts and is sent to extrusion Machine die head, and by after mouth die molding, it is thus achieved that the product wanted.In whole extrusion, the process variable pair such as temperature, pressure Product quality is most important, and the fluctuation of temperature can cause the fluctuation of product quality, and for some temperature-sensitive materials, temperature is too high also It is likely to result in material degradation, causes the generation of waste material, so accurate temperature controls most important.But existing extruder Temperature control system, major part makes a search just for heating system, not in view of the control of cooling system.Cooling system The unnecessary heat produced when system is for taking away screw rod rotational shear material.In recent years, some temperature control systems are had to consider The control of cooling system, but it is undesirable to control result, and adding gentle cooling can constantly switch, and wastes the energy.

Summary of the invention

In order to solve above-mentioned technical problem, it is an object of the invention to: a kind of high accuracy based on positive and negative input system is provided And the distribution type generalized forecast control method of good energy-conserving effect.

The technical solution adopted in the present invention is: distribution type generalized forecast control method based on positive and negative input system, bag Include following steps:

A, sensor from positive and negative input system obtain the positive input of this system, reverse pumping enters, positive input parameter and reverse pumping enter Parameter；

B, positive input parameter and reverse pumping enter parameter synthesis is single input parameter；

C, utilize GPC algorithm and single input parameter to enter to process to positive input and reverse pumping, thus obtain Single input；

D, enter the time lag that the equivalence relation between single input, the time lag of positive input and reverse pumping enter according to positive input, reverse pumping The expection positive input or the expection reverse pumping that calculate positive and negative input system enter；

E, enter to align anti-input system be controlled according to expection positive input or expection reverse pumping.

Further, in described step A, positive input parameter includes positive input Delay Parameters θ_h, positive input gain K_hThe most defeated Angle of incidence constant, τ_h, described reverse pumping enters parameter and includes anti-input delay parameter θ_c, anti-input gain K_cWith anti-input time constant τ_c。

Further, in described step B, single input parameter includes single input Delay Parameters θ, single input gain K and single input Timeconstantτ, the computing formula of described single input parameter is respectively $\mathrm{\τ}\frac{{\mathrm{\τ}}_h+{\mathrm{\τ}}_c}{2},K=\frac{K_h+K_c}{2}$ With θ=min (θ_h, θ_c)。

Further, described positive input, reverse pumping enter the equivalence relation between single input and are: corresponding the most defeated of described positive input The total output going out to enter corresponding reinfusion with described reverse pumping should be equal to the list output corresponding to single input.

Further, during described expection positive input and expection reverse pumping enter, at least the value of is zero.

Further, described positive and negative input system is temperature control system, and described positive input is power of heating, and described reverse pumping enters For cooling power.

Further, described positive and negative input system is in PH and control system, and described positive input is alkaline reagent amount, described instead Input is acid reagent amount.

Further, described positive and negative input system is control pressurer system, and described positive input is aeration quantity, described reverse pumping enter for Rate of air sucked in required.

The invention has the beneficial effects as follows: the present invention enters parameter also by reading positive input parameter and reverse pumping during controlling Both are synthesized single input parameter, has considered the positive and negative input in positive and negative input system, then utilized Generalized Prediction control The time lag characteristic of algorithm processed and positive and negative input calculates, and control process is reflected in single input parameter, thus real The energy during controlling or the consumption of material is reduced on the basis of having ensured now control accuracy.

Accompanying drawing explanation

Fig. 1 is the method step flow chart of the present invention；

Fig. 2 dual input single output system schematic diagram；

Fig. 3 is extruder sketch；

Fig. 4 is that AGPC of the present invention controls result；

Fig. 5 is that traditional double SISO GPC controls result；

Fig. 6 is energy expenditure comparison diagram；

Fig. 7 controls result for point journey PID；

Fig. 8 is the control result that AGPC controls for barrel temperature.

Detailed description of the invention

Below in conjunction with the accompanying drawings the detailed description of the invention of the present invention is described further:

With reference to Fig. 1, distribution type generalized forecast control method based on positive and negative input system, include following steps:

A, sensor from positive and negative input system obtain the positive input of this system, reverse pumping enters, positive input parameter and reverse pumping enter Parameter；

B, positive input parameter and reverse pumping enter parameter synthesis is single input parameter；

C, GPC algorithm and single input parameter is utilized to enter to process to positive input and reverse pumping, thus To single input；

D, enter the time lag that the equivalence relation between single input, the time lag of positive input and reverse pumping enter according to positive input, reverse pumping The expection positive input or the expection reverse pumping that calculate positive and negative input system enter；

E, enter to align anti-input system be controlled according to expection positive input or expection reverse pumping.

The process of above A-E step is based primarily upon three step process flow process:

The first step, the positive input and the reverse pumping that obtain positive and negative input system enter, and enter positive input with reverse pumping comprehensively to become single input, Will positive input and counter enter through positive and negative input system produce output be considered as what single input was produced by positive and negative input system Output (preservation of energy)；

Second step, utilizes GPC algorithm to ask for single input variable；

3rd step, determines the expection input value for controlling according to preservation of energy.

Below as a example by temperature control system, the concrete steps flow process of the present invention is described.

The first step, the positive input and the reverse pumping that obtain positive and negative input system enter, and enter positive input with reverse pumping comprehensively to become single input, Will positive input and counter enter through positive and negative input system produce output be considered as what single input was produced by positive and negative input system Output (preservation of energy).Temperature controlled process can represent with Fig. 2: G_hRepresentative is heated u_hAnd the transmission letter between temperature y Number.Similarly, G_cRepresent cooling u_cAnd the transmission function between temperature y.

In general, temperature course can add the transmission function representation of time lag with single order the following:

$y\left(s\right)=\frac{K_h{e}^{-{\mathrm{\θ}}_{h}s}}{{\mathrm{\τ}}_{h}s+1}{u}_h\left(s\right)+\frac{K_c{e}^{-{\mathrm{\θ}}_{c}s}}{{\mathrm{\τ}}_{c}s+1}{u}_c\left(s\right)---\left(1\right)$

Subscript h represents and heats, and c represents cooling, y (s) and u (s) and represents output and input variable respectively, when θ, K, τ represent Stagnant, gain and time constant.In described step A, positive input parameter includes positive input Delay Parameters θ_h, positive input gain K_hWith Positive input timeconstantτ_h, described reverse pumping enters parameter and includes anti-input delay parameter θ_c, anti-input gain K_cFashionable with reverse pumping Between constant, τ_c.Formula (1) can discrete turn to:

$y\left(t\right)=\frac{b_{h0}{z}^{-1}}{1+a_{h1}{z}^{-1}}{u}_h(t-d_h)+\frac{b_{c0}{z}^{-1}}{1+a_{c1}{z}^{-1}}{u}_c(t-d_c)---\left(2\right)$

In the present invention, positive and negative two inputs are regarded as a single input: cooling power as negative input variable (scope for- 100～0)；Power of heating is positive input variable (scope is 0～100), so formula (1) can be following single input with abbreviation Single output (SISO) model:

$y\left(s\right)=\frac{{\mathrm{Ke}}^{-\mathrm{\ρs}}}{\mathrm{\τs}+1}u\left(s\right)---\left(3\right)$

Notice positive input parameter { τ_h,θ_h,K_h(characteristic i.e. heated) and reverse pumping enter parameter { τ_c,θ_c,K_c(i.e. cool down Characteristic) there is larger difference, during temperature controls this single-input single-output, we are chosen as time constant and gain Adding the average of gentle cooling, time lag then selects to add the minima in gentle cooling, it may be assumed that

$\mathrm{\τ}=\frac{{\mathrm{\τ}}_h+{\mathrm{\τ}}_c}{2},K=\frac{K_h+K_c}{2},\mathrm{\θ}=\min ({\mathrm{\θ}}_h,{\mathrm{\θ}}_c)---\left(4\right)$

After discretization, pseudoprocess can be expressed as:

$y\left(t\right)=\frac{b_0{z}^{-1}}{1+a_1{z}^{-1}}u(t-d)---\left(5\right)$

Second step, utilizes GPC algorithm to ask for single input variable；

After Clark proposes generalized predictive control (GPC) algorithm, GPC control algolithm is applied to chemical industry, oil, doctor The fields such as treatment.As a kind of advance control algorithm, its powerful property is to process nonlinear system, time lag system, mixes and be System and the system of belt restraining, so in this invention, we utilize GPC design controller to ask for pseudo-input, defeated through deriving Go out can be expressed as following form:

Δ U=(G^TG+λI)^-1G(w-f)

$\mathrm{\ΔU}=\left[\begin{array}{c}\mathrm{\Δu}\left(t\right)\\ \mathrm{\Δu}(t+1)\\ .\\ .\\ .\\ \mathrm{\Δu}(t+N_u-1)\end{array}\right],G=\left[\begin{array}{cccccc}{g}_0& 0& .& .& .& 0\\ g_1& g_0& .& .& .& 0\\ .& .& & .& & .\\ .& .& & .& & .\\ .& .& & .& & .\\ g_{N-1}& g_{N-2}& .& .& .& g_{N-N_u}\end{array}\right]---\left(6\right)$

F represents free response, and λ is control matrix, and N is prediction step, N_uBeing to control step-length, w is prediction output, in order to Output y (t) making current time the most smoothly reaches setting value, generally the following first-order filtering equation of selection:

W (t)=y (t)

W (t+k)=α w (t+k-1)+(1-α) r (t+k), k=1......N₂ (7)

Wherein 0≤α ＜ 1

So output can be expressed as:

$\left[\begin{array}{c}u\left(t\right)\\ u(t+1)\\ .\\ .\\ .\\ u(t+N_u-1)\end{array}\right]=\left[\begin{array}{cccccc}1& 0& .& .& .& 0\\ 1& 1& .& .& .& 0\\ .& .& & .& & .\\ .& .& & .& & .\\ .& .& & .& & .\\ 1& 1& .& .& .& 1\end{array}\right]\left[\begin{array}{c}\mathrm{\Δu}\left(t\right)\\ \mathrm{\Δu}(t+1)\\ .\\ .\\ .\\ \mathrm{\Δu}(t+N_u-1)\end{array}\right]+u(t-1)---\left(8\right)$

3rd step, determines the expection input value for controlling according to preservation of energy.

The most described positive input, reverse pumping enter the equivalence relation between single input: positive output corresponding to described positive input with Described reverse pumping enters total output of the reinfusion of correspondence should be equal to the list output corresponding to single input.I.e. add in temperature control system The gross energy that the gross energy that gentle cooling consumes should calculate equal to pseudo-model.

2. further for saving energy or material, described expection positive input and expection reverse pumping enter at least value of Be zero, i.e. the sampling instant of each in temperature control system can only carry out adding the one in temperature control and cooling control.

The concrete calculating process of step D of the present invention is described below in conjunction with example, according to positive input, reverse pumping enter with single input it Between the time lag that enters of equivalence relation (the most above-mentioned 1. in energy conservation relation), the time lag of positive input and reverse pumping calculate positive and negative Expection positive input or the expection reverse pumping of input system enter.

According to above-mentioned 1. in energy conservation relation, the output calculated according to formula (5) should calculate equal to by formula (2) The output obtained, therefore draws the relational expression of preservation of energy:

$\frac{b_0{z}^{-1}}{1+a_1{z}^{-1}}u(t-d)=\frac{b_{h0}{z}^{-1}}{1+a_{h1}{z}^{-1}}{u}_h(t-d_h)+\frac{b_{c0}{z}^{-1}}{1+a_{c1}{z}^{-1}}{u}_c(t-d_c)---\left(9\right)$

Time lag d of positive input considered below_hTime lag d entered with reverse pumping_c:

Situation one: d_h=d_c

Here d=d_h=d_c, z is multiplied by formula (9) both sides^d(1+a₁z^-1)(1+a_h1z^-1)(1+a_c1z^-1), and can after displacement Obtain:

b_h0u_h(t)+b_c0u_c(t)=b₀(1+a_h1z^-1)(1+a_c1z^-1)u(t)

-b_h0(a₁z^-1+a_c1z^-1+a₁a_c1z^-2)u_h(t) (10)

-b_c0(a₁z^-1+a_h1z^-1+a₁a_h1z^-2) u_c(t)

The value of the variable on formula (10) the right is all known, can be therefore a constant C by abbreviation on the right of equation.

b_h0u_h(t)+b_c0u_c(t)=C (11)

Here b_h0＞ 0, b_c0＞ 0, u_h(t)∈[0,100],u_cT () ∈ [-100,0], according to rule (b), u_h,u_cCan count It is:

$u_h\left(t\right)=\left\{\begin{array}{c}\min (\frac{C}{b_{h0}},100),\mathrm{ifC}>0\\ 0,\mathrm{otherwise}\end{array}\right.,u_c\left(t\right)=\left\{\begin{array}{c}\min (-100,\frac{C}{b_{c0}}),\mathrm{ifC}<0\\ 0,\mathrm{otherwise}\end{array}\right.---\left(12\right)$

As seen from formula (12), as C > 0 time, it is believed that be system add thermogenetic effect, therefore u_c(t)=0；Same C < when 0, u_h(t)=0。

Situation two: d_h＞ d_c

It is similar to the situation of situation one, rule-based (a), here d=d_c, z is all multiplied by formula (9) both sides^j(1+a₁z^-1)(1 +a_h1z^-1)(1+a_c1z^-1), j=d_c,d_c+1,...,d_h, and define following formula:

b₀(1+a_h1z^-1)(1+a_c1z^-1)=trs_a

b_h0(1+a₁z^-1)(1+a_c1z^-1)=trs_bh (13)

b_c0(1+a₁z^-1) (1+a_h1z^-1)=trs_bc

Formula (9) abbreviation is:

Trs_a × u (t)=trs_bh × u_h(t-d_h+d_c)+trs_bc×u_c(t)

Trs_a × u (t+1)=trs_bh × u_h(t-d_h+d_c+1)+trs_bc×u_c(t+1) (14)

trs_a×u(t+d_h-d_c)=trs_bh × u_h(t)+trs_bc×u_c(t+d_h-d_c)

According to discrete model (2), and model (2) is multiplied by z^j, j=d_c,d_c+1,...,d_h

$\left\{\begin{array}{c}y(t+d_c)=\frac{b_{h0}{z}^{-1}}{1+a_{h1}{z}^{-1}}{u}_h(t-d_h+d_c)+\frac{b_{c0}{z}^{-1}}{1+a_{c1}{z}^{-1}}{u}_c\left(t\right)\\ y(t+d_c+1)=\frac{b_{h0}{z}^{-1}}{1+a_{h1}{z}^{-1}}{u}_h(t-d_h+d_c+1)+\frac{b_{c0}{z}^{-1}}{1+a_{c1}{z}^{-1}}{u}_c(t+1)\\ ......\\ y(t+d_h)=\frac{b_{h0}{z}^{-1}}{1+a_{h1}{z}^{-1}}{u}_h\left(t\right)+\frac{b_{c0}{z}^{-1}}{1+a_{c1}{z}^{-1}}{u}_c(t+d_h-d_c)\end{array}\right.---\left(15\right)$

By above-mentioned formula it can be seen that y (t+d_c) by up-to-date u_h(t-d_h+d_c) and u_cT () determines, according to rule (b), right In a known temperature, heating or cooling down to have a use.u_h(t-d_h+d_c) it is known that according to formula (14) and (15), therefore u_cT () can be calculated by following formula:

$u_c\left(t\right)=\left\{\begin{array}{c}\frac{\mathrm{trs}\_a\&times;u\left(t\right)-\mathrm{trs}\_\mathrm{bh}\&times;u_h(t-d_h+d_c)-b_{c0}(a_1+a_{h1}+a_1{a}_{h1}{z}^{-1})u_c(t-1)}{b_{c0}}\\ \mathrm{if}{u}_h(t-d_h+d_c)=0\\ 0,\mathrm{if}{u}_h(t-d_h+d_c)>0\end{array}\right.---\left(16\right)$

Then according to result of calculation u_cT () constrains in the range of [-100,0].Same can obtain u_c(t+1),u_c (t+2),...,u_c(t+d_h-d_c-1)。u_h(t) can by formula (15) last formula obtain, the result above asked for And the result (8) that GPC calculates, rearrange formula, it is possible to obtain:

b_h0u_h(t)+b_c0u_c(t+d_h-d_c)=

b₀(1+a_h1z^-1)(1+a_c1z^-1)×u(t+d_h-d_c)

-b_h0(a₁+a_c1+a₁a_c1z^-1) × u_h(t-1) (17)

-b_c0(a₁+a_h1+a₁a_h1z^-1) × u_c(t+d_h-d_c-1)

Substituting into known quantity, formula (17) the right abbreviation is C₁

b_h0u_h(t)+b_c0u_c(t+d_h-d_c)=C₁ (18)

Follow the principles (b), it is possible to obtain result below:

$u_h\left(t\right)=\left\{\begin{array}{c}\min (\frac{C_1}{b_{h0}},100),\mathrm{if}{u}_c\left(t\right)=0\mathrm{and}{C}_1>0\\ 0.\mathrm{if}{u}_c\left(t\right)\&NotEqual;0\mathrm{or}{C}_1\&le;0\end{array}\right.---\left(19\right)$

Situation three: d_h＜ d_c

In this case, computational methods are with situation 2, and its result of calculation is as follows:

$u_h\left(t\right)=\left\{\begin{array}{c}\frac{\mathrm{trs}\_a\&times;u\left(t\right)-\mathrm{trs}\_\mathrm{bc}\&times;u_c(t-d_c+d_h)-b_{h0}(a_1+a_{c1}+a_1{a}_{c1}{z}^{-1})u_h(t-1)}{b_{h0}}\\ \mathrm{if}{u}_c(t-d_c+d_h)=0\\ 0,\mathrm{if}{u}_c(t-d_c+d_h)>0\end{array}\right.---\left(20\right)$

C₂=b₀(1+a_c1z^-1)(1+a_h1z^-1)×u(t+d_c-d_h)-b_c0(a₁+a_h1+a₁a_h1z^-1) × u_c(t-1)

Wherein-b_h0(a₁+a_c1+a₁a_c1z^-1) × u_h(t+d_c-d_h-1)

Being further used as preferred embodiment, described positive and negative input system is temperature control system, and described positive input is Heating power, described reverse pumping enters for cooling power.

In order to check the effectiveness of this method, it is compared with traditional double SISO GPC algorithms:

I, emulation are compared

Energy expenditure compares:

Their parameter selects: prediction step=35, control step-length=8, control weighting=20, and smoothing factor α= 0.95, with reference to Fig. 4 and Fig. 5, can be drawn by controlling result, two control control method can tracking fixed valure and suppressing Interference.Total energy consumption curve represents in figure 6, and this figure can be clearly seen that AGPC algorithm of the present invention is the most prominent energy-conservation Go out.By AGPC algorithm, efficiency consumption decreases at least 65%.

Control results contrast:

Point journey pid algorithm most like with the inventive method compares in control result, and the parameter of point journey PID selects logical Cross trial and error method.Parameter is chosen as, P=3;I=0.01;D=0, a point journey rule is:

$u_h\left(t\right)=\left\{\begin{array}{c}u\left(t\right),\mathrm{ifu}\left(t\right)>0\\ 0,\mathrm{otherwise}\end{array}\right.,u_c\left(t\right)=\left\{\begin{array}{c}u\left(t\right),\mathrm{ifu}\left(t\right)<0\\ 0,\mathrm{otherwise}\end{array}\right.---\left(22\right)$

With reference to Fig. 7 and Fig. 2, point journey PID controls the control result of result and the present invention and compares, and the overshoot dividing journey PID is big A lot.This is because, the characteristic adding two processes of gentle cooling is the most different, and the parameter meeting the process of heating not necessarily meets The parameter of cooling procedure.

II, experimental verification

As a example by extruder temperature control, temperature is one of essential condition of being carried out of extrusion.Material is in extrusion During the source of heat mainly by two, one is the heat provided by barrel external heat circle；Two is between material and barrel Shearing, fricative heat；The regulation of temperature can only be entered by the heating and cooling system on extruder barrel and control system Row.

With reference to Fig. 3, the barrel of this extruder is divided into five sections, including die head 1, flange the 2, first barrel the 3, second barrel 4 With the 3rd barrel 5.First barrel 3 controls result for the temperature of demonstration the inventive method.

Desired temperature is 190 ° of c, when temperature enters stable state when, bears to one and steps to 150 ° of c.The ginseng of controller Number is chosen as: prediction step=35, controls step-length=8, controls weighting=20, and smoothing factor α=0.95, with reference to Fig. 8, from controlling result On can draw utilization the inventive method, not only adding gentle cooling will not be carried out simultaneously, and controls dry straight.

Being further used as preferred embodiment, described positive and negative input system is in PH and control system, described positive input For alkaline reagent amount, described reverse pumping enters for acid reagent amount.

Being further used as preferred embodiment, described positive and negative input system is control pressurer system, and described positive input is Aeration quantity, described reverse pumping enters for rate of air sucked in required.

It is above the preferably enforcement of the present invention is illustrated, but the invention is not limited to described enforcement Example, those of ordinary skill in the art can also make all equivalents on the premise of spirit of the present invention or replace Changing, deformation or the replacement of these equivalents are all contained in the application claim limited range.

Claims (6)

1. distribution type generalized forecast control method based on positive and negative input system, it is characterised in that: include following steps:

A, sensor from positive and negative input system obtain the positive input of this system, reverse pumping enters, positive input parameter and reverse pumping enter parameter；

B, positive input parameter and reverse pumping enter parameter synthesis is single input parameter；

C, utilize GPC algorithm and single input parameter to enter to process to positive input and reverse pumping, thus obtain single the most defeated Enter；

D, enter the time lag that the equivalence relation between single input, the time lag of positive input and reverse pumping enter according to positive input, reverse pumping and calculate The expection positive input or the expection reverse pumping that go out positive and negative input system enter；

E, enter to align anti-input system be controlled according to expection positive input or expection reverse pumping；

In described step A, positive input parameter includes positive input Delay Parameters θ_h, positive input gain K_hWith positive input time constant τ_h, described reverse pumping enters parameter and includes anti-input delay parameter θ_c, anti-input gain K_cWith reverse pumping angle of incidence constant, τ_c；

In described step B, single input parameter includes single input Delay Parameters θ, single input gain K and single input timeconstantτ, The computing formula of described single input parameter is respectively。

Distribution type generalized forecast control method based on positive and negative input system the most according to claim 1, it is characterised in that: Described positive input, reverse pumping enter the equivalence relation between single input: positive output corresponding to described positive input enters with described reverse pumping Total output of corresponding reinfusion should be equal to the list output corresponding to single input.

Distribution type generalized forecast control method based on positive and negative input system the most according to claim 1, it is characterised in that: Described expection positive input and expection reverse pumping enter at least the value of be zero.

Distribution type generalized forecast control method based on positive and negative input system the most according to claim 1, it is characterised in that: Described positive and negative input system is temperature control system, and described positive input is power of heating, and described reverse pumping enters for cooling power.

Distribution type generalized forecast control method based on positive and negative input system the most according to claim 1, it is characterised in that: Described positive and negative input system is in PH and control system, and described positive input is alkaline reagent amount, and described reverse pumping enters for acid reagent Amount.

Distribution type generalized forecast control method based on positive and negative input system the most according to claim 1, it is characterised in that: Described positive and negative input system is control pressurer system, and described positive input is aeration quantity, and described reverse pumping enters for rate of air sucked in required.