CN103941782B - A kind of humiture advanced control method that is applied to warmhouse booth - Google Patents

Abstract

The present invention relates to a kind of advanced control method of the humiture that is applied to warmhouse booth, it is characterized in that: the method is set up green house temperature-humidity model in the SIMULINK of MATLAB, in OPTO22, set up master controller and the diagonal angle decoupling controller based on predictive PI algorithm, connect again MATLAB and OPTO22 and carry out communication, form a diagonal angle decoupling zero control model with predictive PI control algolithm. Based on this special object of warmhouse booth, adopt advanced its temperature and humidity controller of predictive PI control algorithm design, avoid that traditional PID control algorithm governing speed is slow, large problem fluctuates; Noise immunity is strong, and way of realization is simple, controls effectively, can be used for actual industrial process.

Description

A kind of humiture advanced control method that is applied to warmhouse booth

Technical field

The present invention relates to a kind of humiture advanced control method that is applied to warmhouse booth, belong to process control technology neckTerritory.

Background technology

The main application of warmhouse booth is that one can change plant growth environment, according to the optimum growh bar of plant growthPart, regulates greenhouse climate to make it to meet throughout the year plant growth needs, and the not impact of climate and edaphic condition, can avoidExtraneous Four seasons change and the place of harsh weather on its impact, and can on limited soil, produce to the anniversary variousA kind of chamber facility of the anti-season such as vegetables, fresh flower crop.

With other process comparison of industry, warmhouse booth is a typical multiple-input and multiple-output (MIMO) complex nonlinear systemSystem, each envirment factor skewness greenhouse in, the characteristic that there is non-linear, time variation, greatly inertia and large time delay and intercouple,The control of each envirment factor is also also not exclusively independent, and control loop is coupled mutually. When temperature raises, humidity is fallenLow, when humidity increases, temperature will decline, and when a factor is controlled by given setting value, always causes another factorVariation. Therefore the Temperature and Humidity Control of warmhouse booth uses simple pid control algorithm to be difficult to reach the order that closed loop is controlled automatically, more cannot ensure control accuracy and runtime, be difficult to the control effect that reaches satisfied.

Traditional greenhouse control system adopts single-factor control conventionally, and control mode is relatively simple, but ignored environment because ofCoupling between son, does not consider variation and the impact of other factors while regulating a certain envirment factor. As do not examine while controlling temperatureConsider humidity and change the impact on temperature, and generally only control single thermoregulation mechanism, do not consider other executing agency's actionsOn the impact of temperature, such a single-factor control is difficult to ensure multiple-input and multiple-output (MIMO) coupling warmhouse booth control systemThere is good control effect. What a large amount of modern greenhouse control system adopted is all that multiple-factor is coordinated to control, employing be expertEmpirical fuzzy control strategy, sets up fuzzy control rule by expertise, thereby each envirment factor is controlled in comprehensive coordination, thisKind of control mode is more effective to switching mode control object, but control accuracy is not very high, cannot precisely controlled amount andControl law lacks certain continuity.

All there is limitation separately in the temperature/humidity control method of above-mentioned warmhouse booth, is difficult to reach gratifying controlEffect.

Summary of the invention

The object of this invention is to provide a kind of close coupling and the large time delay and anti-that can eliminate humiture in warmhouse boothInterference is strong, meets the advanced control method of its application.

In order to achieve the above object, technical scheme of the present invention has been to provide a kind of humiture that is applied to warmhouse booth firstEnter control method, it is characterized in that, step is:

Step 1, set up warmhouse booth humiture advanced control system, this system comprises that the temperature based on predictive PI algorithm is pre-Survey PI controller and humidity PI controller, decoupling controller and green house temperature-humidity process model based on diagonal angle decoupling zero, peopleFor given temperature input signal SP1 (t) and the humidity input signal SP2 (t) that needs booth to reach, calculate temperature system mistakePoor Error1 (t) and humidity system error E rror2 (t), Error1 (t)=SP1 (t)-PV1 (t), Error2 (t)=SP2 (t)-PV2 (t), PV1 (t) and PV2 (t) are respectively output temperature and the output humidity of warmhouse booth humiture advanced control system,PV1 (t) and PV2 (t) are respectively real time temperature and the real-time humidity of warmhouse booth;

Step 2, by temperature system error E rror1 (t) and humidity system error E rror2 (t) respectively as temperature predictionThe input of PI controller and humidity PI controller, obtains temperature output signal OP1 (t) and humidity output signal OP2 (t),The input/output relation of temperature prediction PI controller or humidity PI controller is:

$u\left(t\right)=K(1+\frac{1}{p{T}_i})e\left(t\right)-\frac{1}{p{T}_i}[u\left(t\right)-u(t-L)],$ Wherein, p is differential operator, and e (t), u (t) are respectively temperatureThe input and output of degree predictive PI controller or humidity PI controller, the inverse that K is process gain, T_iFor control procedure masterLead time constant, L is control procedure lag time;

The temperature transfer function D of step 3, design decoupling controller₁₁(s), temperature is to humidity effect transfer function D₁₂(s)、Humidity affects transfer function D21 (s) and humidity transfer function D to temperature₂₂(s), wherein:

D₁₁(s)=e^-4s； $D_{12}\left(s\right)=\frac{8.2s+1}{3(\mathrm{10,1}s+1)};D_{21}\left(s\right)=\frac{1.1(9.2s+1)}{1.8(7.6s+1)};$ D₂₂(s)=e^-5s；

By temperature output signal OP1 (t) as temperature transfer function D₁₁(s) and humidity temperature is affected to transfer function D₂₁(s) input, by humidity output signal OP2 (t) as temperature to humidity effect transfer function D₁₂And humidity transfer function (s)D₂₂(s) input, by temperature transfer function D₁₁(s) output and temperature are to humidity effect transfer function D₁₂(s) output is rolled upThe long-pending temperature signal MV1 (t) obtaining after decoupling zero, affects transfer function D by humidity to temperature₂₁(s) output and humidity are transmitted letterNumber D₂₂(s) output is done convolution and is obtained the moisture signal MV2 (t) after decoupling zero;

Step 4, set up greenhouse temperature control object process model G_p11(s), greenhouse temperature is to humidity effect control object mistakeJourney model G_p12(s), booth humidity affects control object process model G to temperature_p21And booth humidity control object process mould (s)Type G_p22(s), wherein:

G_p11(s) transfer function model is:

G_p12(s) transfer function model is:

G_p21(s) transfer function model is:

G_p22(s) transfer function model is:

By temperature signal MV1 (t) as greenhouse temperature control object process model G_p11(s) and temperature affect control objectProcess model G_p21(s) input, by moisture signal MV2 (t) as greenhouse temperature to humidity effect control object process modelG_p12And booth humidity control object process model G (s)_p22(s) input, by greenhouse temperature control object process model G_p11(s)Output and greenhouse temperature to humidity effect control object process model G_p12(s) output is done convolution and is obtained output temperature PV1(t), booth humidity is affected to control object process model G to temperature_p21(s) output and booth humidity control object process mouldType G_p22(s) output is done convolution and is obtained exporting humidity PV2 (t), and output temperature PV1 (t) and output humidity PV2 (t) are fed back toTemperature prediction PI controller and humidity PI controller, finally make itself and temperature input signal SP1 (t) and humidity input signalThe error of SP2 (t) progressively reduces, and tends towards stability close to 0.

The invention provides a kind of humiture advanced control algorithm that is applied to warmhouse booth, comprise process of establishing model, pre-Survey PI and diagonal angle Uncoupling Control Based, described control method comprises the control method to temperature and the control method to humidity.The present invention is based on one order inertia and add pure Delay Process object, adopt predictive PI control algolithm and diagonal angle decoupling algorithm design system controlDevice processed.

A kind of humiture advanced control method that is applied to warmhouse booth provided by the invention, in the time that there is interference in the external world, shouldControl loop can, in the control Regression stable state of predictive PI controller, keep good operational effect; The required control of the methodParameter is few, and controls parameter explicit physical meaning, is convenient to parameter tuning.

Method provided by the invention has overcome the deficiencies in the prior art, by adopting advanced control algolithm to solve boothIn the situation of humiture close coupling large time delay, can effectively overcome that humiture influences each other in booth and hysteresis quality is largeSituation, strong interference immunity, way of realization is simple, controls effectively, can be used for actual industrial process.

Brief description of the drawings

Fig. 1 a kind of advanced control loop block diagram of humiture that is applied to warmhouse booth provided by the invention;

Fig. 2 predictive PI controller structure chart;

Fig. 3 diagonal angle decoupling controller structure chart;

The advanced control program of humiture of Fig. 4 warmhouse booth;

In Fig. 5 booth, temperature is not adding the trend under predictive PI algorithm and other algorithms while disturbing;

In Fig. 6 booth, humidity is not adding the trend under predictive PI algorithm and other algorithms while disturbing;

In Fig. 7 booth, temperature is adding the trend under predictive PI algorithm and other algorithms while disturbing;

In Fig. 8 booth, humidity is adding the trend under predictive PI algorithm and other algorithms while disturbing.

Detailed description of the invention

For the present invention is become apparent, hereby with preferred embodiment, and coordinate accompanying drawing to be described in detail below.

The invention provides a kind of humiture advanced control method that is applied to warmhouse booth in conjunction with Fig. 1, Fig. 3 and Fig. 4, itsStep is:

Step 1, set up warmhouse booth humiture advanced control system, this system comprises that the temperature based on predictive PI algorithm is pre-Survey PI controller and humidity PI controller, decoupling controller and green house temperature-humidity process model based on diagonal angle decoupling zero, peopleFor given temperature input signal SP1 (t) and the humidity input signal SP2 (t) that needs booth to reach, calculate temperature system mistakePoor Error1 (t) and humidity system error E rror2 (t), Error1 (t)=SP1 (t)-PV1 (t), Error2 (t)=SP2 (t)-PV2 (t), PV1 (t) and PV2 (t) are respectively output temperature and the output humidity of warmhouse booth humiture advanced control system,PV1 (t) and PV2 (t) are respectively real time temperature and the real-time humidity of warmhouse booth;

Step 2, by temperature system error E rror1 (t) and humidity system error E rror2 (t) respectively as temperature predictionThe input of PI controller and humidity PI controller, obtains temperature output signal OP1 (t) and humidity output signal OP2 (t),The input/output relation of temperature prediction PI controller or humidity PI controller is:

$u\left(t\right)=K(1+\frac{1}{p{T}_i})e\left(t\right)-\frac{1}{p{T}_i}[u\left(t\right)-u(t-L)],$ Wherein, p is differential operator, and e (t), u (t) are respectively temperatureThe input and output of degree predictive PI controller or humidity PI controller, the inverse that K is process gain, T_iFor control procedure masterLead time constant, L is control procedure lag time.

In conjunction with Fig. 2, the present invention adopts predictive PI control after to this delay object decoupling zero, and its principle is based on controlled deviceProcess model G_p(s), suppose the closed loop transfer function, G of expectation_o(s) transfer function that, obtains required controller is:

One order inertia adds pure Delay Process model, can use transfer functionRepresent. Suppose expectationClosed loop transfer function, isIn formula: λ is adjustable parameter. In the time of λ=1, the open loop of system and the sound of closed loopIdentical between seasonable; In the time of λ<1, the closed loop response of system is faster than open-loop response; As λ>1 time, the closed loop response of system is louder than open loopShould be slow. The transfer function of controller is:

$G_c\left(s\right)=\frac{G_o\left(s\right)}{G_p\left(s\right)(1-G_o\left(s\right))}=\frac{\mathrm{Ts}+1}{k_p(\mathrm{\&lambda;Ts}+1-e^{-\mathrm{\&tau;s}})}$

The input/output relation of controller is:

$U\left(s\right)=\frac{1}{\mathrm{\&lambda;}{k}_p}(1+\frac{1}{\mathrm{Ts}})E\left(s\right)-\frac{1}{\mathrm{\&lambda;Ts}}(1-e^{-\mathrm{\&tau;s}})U\left(s\right);$

Above formula the right Section 1 has the version of PI controller; Section 2 can be interpreted as: controller exists_tMomentThe prediction of output based on time interval [t-τ, t] when output obtains. Proportionality constant is roughly the inverse of target gain,Be the time constant of process the time of integration, and the parameter of predicted portions is relevant with the lag time of process object and time constant. ThisPlant controller and be called predictive PI controller (PPI).

The temperature transfer function D of step 3, design decoupling controller₁₁(s), temperature is to humidity effect transfer function D₁₂(s)、Humidity affects transfer function D to temperature₂₁And humidity transfer function D (s)₂₂(s), to input, to export data as basis joint portionDivide mechanism, set up " ash bin " model of each object. Wherein:

D₁₁(s)=e^-4s； $D_{12}\left(s\right)=\frac{8.2s+1}{3(\mathrm{10,1}s+1)};D_{21}\left(s\right)=\frac{1.1(9.2s+1)}{1.8(7.6s+1)};$ D₂₂(s)=e^-5s；

By temperature output signal OP1 (t) as temperature transfer function D₁₁(s) and humidity temperature is affected to transfer function D21(s) input, by humidity output signal OP2 (t) as temperature to humidity effect transfer function D₁₂And humidity transfer function (s)D₂₂(s) input, by temperature transfer function D₁₁(s) output and temperature are to humidity effect transfer function D₁₂(s) output is rolled upThe long-pending temperature signal MV1 (t) obtaining after decoupling zero, affects transfer function D by humidity to temperature₂₁(s) output and humidity are transmitted letterNumber D₂₂(s) output is done convolution and is obtained the moisture signal MV2 (t) after decoupling zero.

The present invention adopts dynamic decoupling---diagonal matrix decoupling control method to this multi-variable system. Consider following 2 × 2Multivariable System with Time-delay:

$G_p\left(s\right)=\left\{\begin{array}{cc}{G}_{p11}\left(s\right)& G_{12}\left(s\right)\\ G_{p21}\left(s\right)& G_{p22}\left(s\right)\end{array}\right\}=\left\{\begin{array}{cc}{G}_{11}\left(s\right)e^{-l_{11}\left(s\right)}& G_{12}\left(s\right)e^{-l_{12}\left(s\right)}\\ G_{21}\left(s\right)e^{-l_{21}\left(s\right)}& G_{22}\left(s\right)e^{-l_{22}\left(s\right)}\end{array}\right\}$

If decoupling controller has following version: $D\left(s\right)=\left\{\begin{array}{cc}1& d_{12}\left(s\right)\\ d_{21}\left(s\right)& 1\end{array}\right\}$

, according to Decoupling Conditions, G (s) D (s) is diagonal matrix, has:

$\left\{\begin{array}{c}{G}_{11}\left(s\right)e^{-l_{11}\left(s\right)}{d}_{12}\left(s\right)+G_{12}\left(s\right)e^{-l_{12}\left(s\right)}=0\\ G_{21}\left(s\right)e^{-l_{21}\left(s\right)}+G_{22}\left(s\right)e^{-l_{22}\left(s\right)}{d}_{21}\left(s\right)=0\end{array}\right.$

Can obtain d by separating above formula₁₂(s)、d₂₁(s)。

Under many circumstances, d₁₂(s)、d₂₁(s) may be not attainable, use a series of method to process,Make it to realize.

Can obtain according to (1)

$D\left(s\right)=\left\{\begin{array}{cc}1& \frac{-G_{12}\left(s\right)}{G_{11}\left(s\right)}{e}^{-(l_{12}-l_{11})s}\\ \frac{-G_{21}\left(s\right)}{G_{22}\left(s\right)}{e}^{-(l_{21}-l_{22})s}& 1\end{array}\right\}$

If (l₁₂-l₁₁)<⁰Or (l₁₂-l₂₂) < 0, D (s) is not for attainable. Carry out as follows to D (s)

Amendment can realize it

$D\left(s\right)=\left\{\begin{array}{cc}{e}^{-w(l_{22}-l_{21})s}& \frac{-G_{12}\left(s\right)}{G_{11}\left(s\right)}{e}^{-w(l_{12}-l_{11})s}\\ \frac{-G_{21}\left(s\right)}{G_{22}\left(s\right)}{e}^{-w(l_{21}-l_{22})s}& e^{-w(l_{11}-l_{12})s}\end{array}\right\}$

Wherein: $w\left(x\right)=\left\{\begin{array}{cc}x,& \mathrm{ifx}\&GreaterEqual;0\\ 0,& \mathrm{ifx}<0\end{array}\right.$

Through processing, D (s) can realize, and G (s) D (s) is also diagonal matrix.

G (s) D (s) is expressed as to diagonal matrix

M(s)=G(s)D(s)=diag(m₁(s)，m₂(s)，…，m_n(s)). After Decoupling design, just can adopt various elder generationsEnter control algolithm for each relatively independent generalized process object m₁(s)，m₂(s)，…，m_n(s) CONTROLLER DESIGN.

Step 4, set up greenhouse temperature control object process model G_p11(s), greenhouse temperature is to humidity effect control object mistakeJourney model G_p12(s), booth humidity affects control object process model G to temperature_p21And booth humidity control object process mould (s)Type G_p22(s), wherein:

G_p11(s) transfer function model is:

G_p12(s) transfer function model is:

G_p21(s) transfer function model is:

G_p22(s) transfer function model is:

By temperature signal MV1 (t) as greenhouse temperature control object process model G_p11(s) and temperature affect control objectProcess model G_p21(s) input, by moisture signal MV2 (t) as greenhouse temperature to humidity effect control object process modelG_p12And booth humidity control object process model G (s)_p22(s) input, by greenhouse temperature control object process model G_p11(s)Output and greenhouse temperature to humidity effect control object process model G_p12(s) output is done convolution and is obtained output temperature PV1(t), booth humidity is affected to control object process model G to temperature_p21(s) output and booth humidity control object process mouldType G_p22(s) output is done convolution and is obtained exporting humidity PV2 (t), and output temperature PV1 (t) and output humidity PV2 (t) are fed back toTemperature prediction PI controller and humidity PI controller, finally make itself and temperature input signal SP1 (t) and humidity input signalThe error of SP2 (t) progressively reduces, and tends towards stability close to O.

For realizing the present invention, hardware device is selected OPTO22 controller, and OPTO22 is one of PAC product, by the U.S.Opto22 company manufactures, and it is found in 1974, for opening in industrial automation, remote monitoring and business data collection aspectSend out and manufacture hardware and software application product. By employing standard and commercial internet, networking and computer technology, OPTO22'sI/O and control system, allow client from its business is managed vital all machinery, Electrical and Electronic assetsCarry out data monitoring, control and collection. The products & services of OPTO22 are automation end user, original equipment manufacturer(OEM) and information technology and administrative staff provide support.

The present invention has developed the Real-Time Monitoring software based on OPTO22PACProjectProfessional.PACProjectProfessional comprises following a few part composition: PACControl, PACDisplay, OPTOOPCServer, PACManager etc. Algorithm adopts PACControl script to programme, and user monitoring interface adoptsPACDisplay carries out interface development, the steps include:

Step 1: in the SIMULINK of MATLAB, set up the humiture process model of warmhouse booth, as shown in table 1;

Step 2: in OPTO22, set up shown in the master controller based on predictive PI algorithm shown in Fig. 2 and Fig. 3 based onThe decoupling controller of diagonal angle decoupling zero, its algorithm principle sees above, and transfer function is in shown in table 2 and table 3;

Step 3: connect MATLAB and OPTO22 and carry out communication, make master controller and decoupling controller, warmhouse booth warm and humidDegree is controlled model and is connected and composed successively close loop negative feedback control model.

In above control system, in fact realize following two step functions (as shown in Figure 4):

The first step, is used diagonal angle decoupling controller to overcome the intercouple characteristic of impact of humiture in booth;

Second step, uses predictive PI controller to complete the control respectively to temperature in booth and humidity, can obtain satisfiedControl effect and strong interference immunity.

Fig. 5 and Fig. 6 are the systematic steady state temperature that does not adopt predictive PI controller while not disturbing and adopt predictive PI controllerDegree and the comparison of humidity output waveform. The response speed of predictive PI controller than traditional PID controll block very as seen from the figureMany, play good effect.

Fig. 7 and Fig. 8, while adding a step signal as interference to control system, do not adopt predictive PI controller and adoptWith the systematic steady state temperature and humidity output waveform comparison of predictive PI controller. Can find out the system that adopts predictive PI controllerAntijamming capability is obviously more intense, and it can ensure not cause in the situation of very large overshoot, returns to setting value with speed faster.Prove that a kind of humiture advanced control method that is applied to warmhouse booth that the present invention proposes has good control effect.

Claims (1)

1. a humiture advanced control method that is applied to warmhouse booth, is characterized in that, step is:

Step 1, set up warmhouse booth humiture advanced control system, this system comprises the temperature prediction PI based on predictive PI algorithmController and humidity PI controller, decoupling controller and green house temperature-humidity process model based on diagonal angle decoupling zero, artificially giveSurely the temperature input signal SP1 (t) and the humidity input signal SP2 (t) that need booth to reach, calculate temperature system errorError1 (t) and humidity system error E rror2 (t), Error1 (t)=SP1 (t)-PV1 (t), Error2 (t)=SP2 (t)-PV2(t), PV1 (t) and PV2 (t) are respectively output temperature and output humidity, the i.e. PV1 of warmhouse booth humiture advanced control system(t) and PV2 (t) be respectively the real time temperature of warmhouse booth and real-time humidity;

Step 2, temperature system error E rror1 (t) and humidity system error E rror2 (t) are controlled as temperature prediction PI respectivelyThe input of device processed and humidity PI controller, obtains temperature output signal 0P1 (t) and humidity output signal 0P2 (t), temperatureThe input/output relation of predictive PI controller or humidity PI controller is:

$u\left(t\right)=K(1+\frac{1}{p{T}_i})e\left(t\right)-\frac{1}{p{T}_i}[u\left(t\right)-u(t-L)],$ Wherein, p is differential operator, and e (t), that u (t) is respectively temperature is pre-Survey the input and output of PI controller or humidity PI controller, the inverse that K is process gain, T_iWhile dominating for control procedureBetween constant, L is control procedure lag time;

The temperature transfer function D of step 3, design decoupling controller₁₁(s), temperature is to humidity effect transfer function D₁₂(s), humidityTemperature is affected to transfer function D₂₁And humidity transfer function D (s)₂₂(s), wherein:

D₁₁(s)=e^-4s； $D_{12}\left(s\right)=\frac{8.2s+1}{3(\mathrm{10,1}s+1)};D_{21}\left(s\right)=\frac{1.1(9.2s+1)}{1.8(7.6s+1)};$ D₂₂(s)=e^-5s；

By temperature output signal OP1 (t) as temperature transfer function D₁₁(s) and humidity temperature is affected to transfer function D₂₁(s)Input, by humidity output signal OP2 (t) as temperature to humidity effect transfer function D₁₂And humidity transfer function D (s)₂₂(s)Input, by temperature transfer function D₁₁(s) output and temperature are to humidity effect transfer function D₁₂(s) output is done convolution and is obtainedTo the temperature signal MV1 (t) after decoupling zero, humidity is affected to transfer function D to temperature₂₁(s) output and humidity transfer function D₂₂(s) output is done convolution and is obtained the moisture signal MV2 (t) after decoupling zero;

Step 4, set up greenhouse temperature control object process model G_p11(s), greenhouse temperature is to humidity effect control object process mouldType G_p12(s), booth humidity affects control object process model G to temperature_p21And booth humidity control object process model (s)G_p22(s), wherein:

G_p11(s) transfer function model is:

G_p12(s) transfer function model is:

G_p21(s) transfer function model is:

G_p22(s) transfer function model is:

By temperature signal MV1 (t) as greenhouse temperature control object process model G_p11(s) and temperature affect control object processModel G_p21(s) input, by moisture signal MV2 (t) as greenhouse temperature to humidity effect control object process model G_p12(s)And booth humidity control object process model G_p22(s) input, by greenhouse temperature control object process model G_p11(s) outputWith greenhouse temperature to humidity effect control object process model G_p12(s) output is done convolution and is obtained output temperature PV1 (t), will be largeCanopy humidity affects control object process model G to temperature_p21(s) output and booth humidity control object process model G_p22(s)Output do convolution and obtain exporting humidity PV2 (t), output temperature PV1 (t) and output humidity PV2 (t) are fed back to temperature predictionPI controller and humidity PI controller, finally make itself and temperature input signal SP1 (t) and humidity input signal SP2 (t)Error progressively reduces, and tends towards stability close to 0.