CN107037728A - Greenhouse optimal control method based on multiple objective gray particle cluster algorithm - Google Patents
Greenhouse optimal control method based on multiple objective gray particle cluster algorithm Download PDFInfo
- Publication number
- CN107037728A CN107037728A CN201710173446.8A CN201710173446A CN107037728A CN 107037728 A CN107037728 A CN 107037728A CN 201710173446 A CN201710173446 A CN 201710173446A CN 107037728 A CN107037728 A CN 107037728A
- Authority
- CN
- China
- Prior art keywords
- greenhouse
- model
- humidity
- particle
- temperature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Greenhouses (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention discloses a kind of greenhouse optimal control method based on multiple objective gray particle cluster algorithm, by introducing artificial governing factor, based on the autoregression model ARX of extension, build temperature, humidity and energy consumption cost multi-objective Model function, on this basis, using grey correlation theory and particle swarm optimization algorithm PSO, multiobjective optimal control is carried out towards greenhouse model.The present invention can make greenhouse save electric cost to a certain extent, and reference is provided for the Reasonable Regulation And Control of greenhouse.
Description
Technical field
The present invention relates to Technique for Controlling Greenhouse Environment field, specifically a kind of temperature based on multiple objective gray particle cluster algorithm
Room environmental optimal control method.
Background technology
Chamber crop production is highly dense type utilities industry, relative to field crop, is not easily susceptible in hothouse production outer
The interference of boundary's weather, can meet the demand to crop in different producers' growth cycle.Agricultural greenhouse is by controller to temperature
Room environmental parameter (such as temperature, humidity, gas concentration lwevel, illumination condition) is regulated and controled, and is at chamber crop growth
Convenient state, and reduce energy resource consumption while improving crop yield, quality and economic benefit.Traditional green house control
Pattern majority depends on the priori of the producer, and subjectivity is strong, poor real, False Rate are high, lacks scientific basis, does not apply to
In the demand of current hothouse production.The key for solving actual Greenhouse System regulation and control problem at present is to study powerful greenhouse
Environment controlling technique, and efficient technology is that based on accurate model structure, therefore, research is adapted to modern greenhouse control
Model structure it is imperative.
The research of current greenhouse flower, which is concentrated mainly on, to be built practical temperature, Humidity Model and chooses effective excellent
Change the aspect of control method two.In terms of greenhouse modeling, mainly there are Vanthoor models, Bennis models, neutral net mould
Type, fuzzy model, DSSAT models etc..In Technique for Controlling Greenhouse Environment research, mainly using various intelligent control methods, such as
PID control, fuzzy control, ANN Control, uneoupled control, multi objective control, evolution algorithm etc..
In the model field towards greenhouse optimal control, Greenhouse System has existence between complicated structure, model
Energy difference, disclosure satisfy that the model of greenhouse flower demand is limited.In greenhouse flower field, conventional PID controller meeting
Occur problems, such as parameter tuning is bad, poor anti jamming capability and to the adaptability of greenhouse it is poor situations such as；Fuzzy Control
The parameter of device processed easily changes, and is difficult to determine fuzzy rule and membership function etc., directly affects the usability of the controller
Energy；Neural network control technique is easily trapped into local minimum point, the low shortcoming of convergence precision in the application.
The content of the invention
It is an object of the invention to provide a kind of greenhouse optimal control method based on multiple objective gray particle cluster algorithm,
To solve the not enough problem of prior art green house control method presence.
In order to achieve the above object, the technical solution adopted in the present invention is：
Greenhouse optimal control method based on multiple objective gray particle cluster algorithm, it is characterised in that：By introducing people
Work governing factor, based on the autoregression model ARX of extension, builds temperature, humidity and energy consumption cost multi-objective Model function,
On this basis, using grey correlation theory and particle swarm optimization algorithm PSO, multiple-objection optimization is carried out towards greenhouse model
Control, specifically includes following steps：
(1), the temperature of multipoint acquisition greenhouse, humidity information, with adaptive weighted Fusion Estimation Algorithm to collection
Temperature, humidity information data Layer carry out data fusion, for greenhouse modeling and optimal control；
(2), by introducing artificial governing factor, based on the autoregression model ARX of extension, with the side of System Discrimination
Method picks out the structure and parameter of model, builds chamber environment temperature, the autoregression model of the extension of humidity, and regulate and control with greenhouse
The electricity consumed in equipment operation is that energy consumption cost model is set up in reference, and detailed process is as follows：
(2.1) chamber environment temperature, the autoregression model of the extension of humidity are set up：
The temperature of greenhouse external environment, humidity, wind speed and intensity of illumination are considered as disturbance input amount, by greenhouse adjusting device wind
Machine, wet curtain, sunshade, spray, roof window are considered as decision-making input quantity, and output variable is the temperature and humidity in greenhouse, and temperature is set up with this
The autoregression model structure of the extension of room environmental epidemic disaster, the form such as following equation of the autoregression model of the extension of epidemic disaster
It is shown：
In formula (1)-(3), y=[y1 y2]TIt is the temperature and humidity in greenhouse；X=[x1 x2…..x9]TFor model
Input quantity；x1For outdoor temperature, unit for DEG C；x2For outside relative humidity；x3For outdoor wind speed, unit is ms-1；x4For light
According to intensity, unit is lux；x5To x9Respectively greenhouse adjusting device blower fan, wet curtain, sunshade, spray, roof window；A(z-1)、B(z-1)
For ARX model coefficient polynomial, wherein A (z-1) it is the corresponding parametric polynomial of output variable, Bi(z-1) it is respectively nine inputs
The corresponding parametric polynomial of variable[12]；z-1For backward shift operator；K is time variable, and unit is min；V (k) is random noise；na
For A (z-1) rank；nb1,nb2.....nb9For B1(z-1)、B2(z-1).....B9(z-1) rank[12]；Because green-house scale is small, it is considered to v
(k) for significantly or be difficult to guess, then formula (1) can be reduced to：
A(z-1)*yi(t)=B (z-1)*xj(t) (i=1,2；J=1,2,3.....9) (4),
According to actual greenhouse situation, highest order is arranged to second order, and bring formula (2), (3) into formula (4), you can
Release greenhouse epidemic disaster expression be：
Known z-1* u (t)=u (t-1), z-2* u (t)=u (t-2), greenhouse ring can be tried to achieve by being substituted into (5) and (6) formula
The autoregression model function expression of the extension of border epidemic disaster；
(2.2) greenhouse energy consumption cost model is set up：
In green house control equipment running process, the electricity of adjusting device consumption is made up of two parts, and a part is continuously may be used
The electricity of equipment operation a period of time consumption is controlled, another part is the electricity that discontinuous controllable device fully opens consumption, energy consumption
Cost model J can be expressed as：
In formula (7),The electricity of consumption is opened for the regulating and controlling mechanism of common n1 discontinuous actions；For
The electricity of common n2 continuous action regulating and controlling mechanism consumption；piFor the rated power of single continuous regulating and controlling mechanism；qiTo be single discontinuous
The rated power of regulating and controlling mechanism；xiFor the on off state of regulating and controlling mechanism；taiFor the full opening of operation of noncontinuity regulating and controlling mechanism
Time；tbiFor continuity regulating and controlling mechanism run time；
(3), by introducing grey correlation theory, in standard grey relevance theory and particle swarm optimization algorithm PSO base
On plinth, adjusting device combination variety is considered as to the solution of particle, using temperature model, Humidity Model and energy consumption model as object function,
The multiobjective optimal control of greenhouse flower is completed with this, detailed process is as follows：
(3.1), temperature, humidity and financial cost are obtained respectively with grey correlation theory and particle swarm optimization algorithm PSO
Adaptive optimal control value of the object function under the constraint of control device combination variety, the sequence being made up of the optimal value of each object function is made
On the basis of vector sequence；
(3.2), according to greenhouse actual conditions, using five kinds of adjusting device permutation and combination species as the solution of particle, in combination
Under the constraints of species, each particle scale, iterations, position and speed etc. are initialized, by the initial value of each particle solution
Change into binary matrix form and substitute into each object function, obtain initial target value, regard this desired value as target vector sequence；
(3.3) degree of association between the base vector sequence and target vector sequence of each particle, is evaluated, by current goal
The individual extreme value that vector sequence degree of association corresponding with its is set to particle is stored as pbest, by the position of the maximum particle of the degree of association
Put with its degree of association and to be stored as gbest as the global extremum of whole population；
(3.4) particle position, speed, Studying factors, weights, are updated, when updating particle position, particle position is carried out
Floor operation；
(3.5), calculate the degree of association of each particle, by its degree of association with after desired positions corresponding to the degree of association make
Compare, if more excellent, current degree of association particle position corresponding with its is stored as pbest；Compare the individual pole of all particles
Value and global extremum, if more excellent, are updated to global extremum, and store its corresponding particle position by the individual extreme value；
(3.6), judge whether current iteration number of times meets stop condition, if 4. be unsatisfactory for return to step continues iteration；It is no
Then, stop search, the corresponding switchgear combination of output global extremum and temperature, humidity and financial cost value.
The present invention is on the basis of fully investigation model method, using autoregression model (the Auto Regressive of extension
EXogenous, ARX) structure, input and output information founding mathematical models using system are not only suitable for linear system, fitted again
For nonlinear system.Because humiture change procedure is non-linear, it can approximately be considered as linear, utilization at humiture equalization point
ARX model being capable of approximate representation.
The present invention passes through the multipoint acquisition and data anastomosing algorithm of sensor, it is ensured that the uniformity of collection information and credible
Degree.Because of the complexity of greenhouse, when gathering greenhouse data, it is larger that the position difference of collection point can cause information to have
Difference.Traditional greenhouse acquisition method majority is acquired to single-point, ignores the otherness of different acquisition point information.Cause
This, study multisensor data acquisition technology, and use adaptive weighted Fusion Estimation Algorithm, it is ensured that information gathering it is consistent
Property and confidence level be one of key issue that the present invention needs to solve.
Greenhouse adjusting device is added humiture model structure, and set up energy by the present invention by introducing artificial governing factor
Source consumption models.Traditional environmental control of greenhouse model majority considers temperature and humidity model structure, ignores the energy of greenhouse
Source consuming cost, to a certain extent and is unsatisfactory for the optimum control of greenhouse.Therefore, the present invention taking into full account temperature, it is wet
On the basis of degree, energy resource consumption model is set up, requirement of the plant growth to environment can be met, electric cost has been saved again.
The present invention surrounds grey correlation theory, with reference to grey correlation theory and particle swarm optimization algorithm (Particle
Swarm Optimization, PSO), complete the multiobjective optimal control of greenhouse flower.Multiple target PSO is used for engineering
Design, chemical field, field of power and Data Mining etc., can effectively solve the problem that the optimization problem of multiple target, in temperature
Research in the control of room environmental system is limited.The present invention is applied to greenhouse on the basis of fully research multiple target PSO
In control, the optimal control of population is completed through successive ignition.
The present invention regulates and controls in Optimal Control Strategy of the research about greenhouse with multiple objective gray PSO to greenhouse
Equipment is optimized, and has certain advantage compared with linear weight sum method, single goal PSO algorithms, can meet greenhouse work
Thing grows the requirement to environment, and electric cost has been saved to a certain extent, reference is provided for the Reasonable Regulation And Control of greenhouse.
Compared with prior art, what advantage is the present invention have：
(1) multipoint acquisition humiture information, improves the accuracy of data.
(2) multiple-objection optimization thought is surrounded, temperature, humidity, energy consumption cost model is set up, meets the growth bar of crop suitably
While part, certain energy consumption cost is saved.
(3) there is discontinuous problem for greenhouse adjusting device executing agency, propose discrete particle cluster algorithm, solve PSO
The problem of shaping parameter can only being optimized.
(4) because of the complexity of greenhouse, there is certain coupling between the indoor multi-source factor, the present invention is to greenhouse ring
When border is regulated and controled, without considering the coupling between information.
Brief description of the drawings
Fig. 1 is the inventive method theory diagram.
Fig. 2 is the output curve diagram in the specific embodiment of the invention with D1 checking model temperatures.
Fig. 3 is the output curve diagram in the specific embodiment of the invention with D1 checking model humidity.
Fig. 4 is the output curve diagram in the specific embodiment of the invention with D2 checking model temperatures.
Fig. 5 is the output curve diagram in the specific embodiment of the invention with D2 checking model humidity.
Embodiment
As shown in figure 1, the greenhouse optimal control method based on multiple objective gray particle cluster algorithm, artificial by introducing
Governing factor, based on the autoregression model (Auto Regressive eXogenous, ARX) of extension, builds temperature, wet
Degree and energy consumption cost multi-objective Model function, on this basis, using grey correlation theory and particle swarm optimization algorithm
(Particle Swarm Optimization, PSO), carries out multiobjective optimal control towards greenhouse model, specifically includes
Following steps：
(1), the temperature of multipoint acquisition greenhouse, humidity information, with adaptive weighted Fusion Estimation Algorithm to collection
Temperature, humidity information data Layer carry out data fusion, for greenhouse modeling and optimal control；The present invention is away from work
The high 1m of thing horizontal zone, temperature-humidity sensor at three is equidistantly spaced from by greenhouse North and South direction, it is ensured that greenhouse information is adopted
The uniformity and confidence level of collection.
(2), by introducing artificial governing factor, based on the autoregression model ARX of extension, with the side of System Discrimination
Method picks out the structure and parameter of model, builds chamber environment temperature, the autoregression model of the extension of humidity, and regulate and control with greenhouse
The electricity consumed in equipment operation is that energy consumption cost model is set up in reference, and detailed process is as follows：
(2.1) chamber environment temperature, the autoregression model of the extension of humidity are set up：
The temperature of greenhouse external environment, humidity, wind speed and intensity of illumination are considered as disturbance input amount, by greenhouse adjusting device wind
Machine, wet curtain, sunshade, spray, roof window are considered as decision-making input quantity, and output variable is the temperature and humidity in greenhouse, and temperature is set up with this
The autoregression model structure of the extension of room environmental epidemic disaster, through cross validation, the humiture curve obtained by model prediction with
The humiture plots changes of actual measurement are consistent.Temperature root-mean-square error RMSE is 0.7 DEG C, and relative error RE is
1.59%.Humidity root-mean-square error RMSE is 1.1%, and relative error RE is 2.14%.Illustrate that ARX model can effectively simulate greenhouse
Temperature and humidity in environment, disclosure satisfy that demand of the environmental control of greenhouse equipment to model structure.
The form of the autoregression model of the extension of epidemic disaster is as shown in following equation：
In formula (1)-(3), y=[y1 y2]TIt is the temperature and humidity in greenhouse；X=[x1 x2…..x9]TFor model
Input quantity；x1For outdoor temperature, unit for DEG C；x2For outside relative humidity；x3For outdoor wind speed, unit is ms-1；x4For light
According to intensity, unit is lux；x5To x9Respectively greenhouse adjusting device blower fan, wet curtain, sunshade, spray, roof window；A(z-1)、B(z-1)
For ARX model coefficient polynomial, wherein A (z-1) it is the corresponding parametric polynomial of output variable, Bi(z-1) it is respectively nine inputs
The corresponding parametric polynomial of variable[12]；z-1For backward shift operator；K is time variable, and unit is min；V (k) is random noise；na
For A (z-1) rank；nb1,nb2.....nb9For B1(z-1)、B2(z-1).....B9(z-1) rank[12]；Because green-house scale is small, it is considered to v
(k) for significantly or be difficult to guess, then formula (1) can be reduced to：
A(z-1)*yi(t)=B (z-1)*xj(t) (i=1,2；J=1,2,3.....9) (4),
According to actual greenhouse situation, highest order is arranged to second order, and bring formula (2), (3) into formula (4), you can
Release greenhouse epidemic disaster expression be：
Known z-1* u (t)=u (t-1), z-2* u (t)=u (t-2), greenhouse ring can be tried to achieve by being substituted into (5) and (6) formula
The autoregression model function expression of the extension of border epidemic disaster；
(2.2) greenhouse energy consumption cost model is set up：
In green house control equipment running process, the electricity of adjusting device consumption is made up of two parts, and a part is continuously may be used
The electricity of equipment operation a period of time consumption is controlled, another part is the electricity that discontinuous controllable device fully opens consumption, energy consumption
Cost model J can be expressed as：
In formula (7),The electricity of consumption is opened for the regulating and controlling mechanism of common n1 discontinuous actions；For
The electricity of common n2 continuous action regulating and controlling mechanism consumption；piFor the rated power of single continuous regulating and controlling mechanism；qiTo be single discontinuous
The rated power of regulating and controlling mechanism；xiFor the on off state of regulating and controlling mechanism；taiFor the full opening of operation of noncontinuity regulating and controlling mechanism
Time；tbiFor continuity regulating and controlling mechanism run time；
(3), by introducing grey correlation theory, in standard grey relevance theory and particle swarm optimization algorithm PSO base
On plinth, adjusting device combination variety is considered as to the solution of particle, using temperature model, Humidity Model and energy consumption model as object function,
The multiobjective optimal control of greenhouse flower is completed with this, detailed process is as follows：
(3.1), temperature, humidity and financial cost are obtained respectively with grey correlation theory and particle swarm optimization algorithm PSO
Adaptive optimal control value of the object function under the constraint of control device combination variety, the sequence being made up of the optimal value of each object function is made
On the basis of vector sequence；
(3.2), according to greenhouse actual conditions, using five kinds of adjusting device permutation and combination species as the solution of particle, in combination
Under the constraints of species, each particle scale, iterations, position and speed etc. are initialized, by the initial value of each particle solution
Change into binary matrix form and substitute into each object function, obtain initial target value, regard this desired value as target vector sequence；
(3.3) degree of association between the base vector sequence and target vector sequence of each particle, is evaluated, by current goal
The individual extreme value that vector sequence degree of association corresponding with its is set to particle is stored as pbest, by the position of the maximum particle of the degree of association
Put with its degree of association and to be stored as gbest as the global extremum of whole population；
(3.4) particle position, speed, Studying factors, weights, are updated, when updating particle position, particle position is carried out
Floor operation；
(3.5), calculate the degree of association of each particle, by its degree of association with after desired positions corresponding to the degree of association make
Compare, if more excellent, current degree of association particle position corresponding with its is stored as pbest；Compare the individual pole of all particles
Value and global extremum, if more excellent, are updated to global extremum, and store its corresponding particle position by the individual extreme value；
(3.6), judge whether current iteration number of times meets stop condition, if 4. be unsatisfactory for return to step continues iteration；It is no
Then, stop search, the corresponding switchgear combination of output global extremum and temperature, humidity and financial cost value.
Specific embodiment：
Step one：Experimental enviroment and data acquisition
Test the tea tree seedling cultivation greenhouse on March 14th, 2016 to April 21 in Agricultural University Of Anhui Nong Cui gardens to carry out, temperature
The glasshouse that room is built by steel construction, ridge is about 80m to for north-south, taking up an area2, skylight, sunshade controllable device are arranged at top, northern
There are 2 Fans in portion, the wet curtain formation wind direction convection current with greenhouse south, is easy to cooling, shower is located at 0.5m above tea shoot.Tea shoot
Size is about 15cm, and the management operation such as cooled to tea shoot according only to artificial experience, humidified in the past.According to Tea Science, expert knows
Know, suitable temperature is 22 DEG C~28 DEG C, humidity is 50%~70%, preferable optimal humiture setting value is 25 DEG C and 60%.
Horizontal zone away from the high 1m of tea shoot, temperature-humidity sensor at three is equidistantly spaced from by North and South direction, with adaptive weighted number
Temperature and humidity is merged in data Layer according to Fusion Estimation Algorithm, for greenhouse modeling and optimal control.Utilize NI
Compact RIO and WSN3202 short-range wireless communication modules gather the temperature of indoor environment, humidity, gas concentration lwevel and
The parameters such as intensity of illumination；With weather informations such as PH automatic weather stations collection outdoor temperature, humidity, wind speed and intensities of illumination.
Step 2：The foundation of model and model checking
The data of collection are pre-processed, melted the warm and humid angle value of multiple spot with adaptive weighting data fusion algorithm
Close, on MATLAB platforms, humiture model structure is set up with ARX model.By analyzing adjusting device, regulation and control are set
It is standby to be divided into continuous regulation and control (blower fan, spray, wet curtain) and discontinuous regulation and control (sunshade, skylight), the electricity that adjusting device is consumed
It is used as energy resource consumption model.Choose March 14 to the data between April 18 and set up greenhouse model, with cross validation
The accuracy of mode testing model.By 19 days 11 April:00~11:30 and 12:00~12:Data during 30 are respectively used to temperature
Humidity modeling, with 19 days 11 April:30~12:00 and 12:30~13:Data during 00 are checking data, are designated as D1With
D2.By D1And D2The input quantity of middle data substitutes into the predicted value that above-mentioned model calculates temperature and humidity respectively, and its result is following such as
Shown in Fig. 2,3,4,5.
Step 3：Algorithm optimization and interpretation of result
On MATLAB platforms, write multiple objective gray particle cluster algorithm and arithmetic result is compared checking.Choose 4
On the moon 19 11:The data of 35 monitorings implement multiobjective optimal control, and the temperature value that the moment monitors is 31.5 DEG C, and humidity value is
47.2%, the adjusting device of unlatching is sunshade, wet curtain and roof window.According to the condition of tea shoot suitable growth, show that greenhouse is stilled need
Cooling operation.Using the temperature set up, humidity, energy consumption model as optimization object function, with blower fan, wet curtain, sunshade, spray and
Roof window is decision variable, and analog simulation is carried out to decision variable and temperature, humidity and energy consumption with multiple objective gray PSO algorithms.
Equipment combination after optimization is sunshade and spray, chooses 11:35 to 11:55 periods carried out analogue simulation to greenhouse：I.e.
In the case where opening combined situation of the sunshade with spray, 11:Indoor temperature is 24.51 DEG C when 55, and humidity is 59.35%, consumed energy
For 0.14J, now the humiture in greenhouse reaches that tea shoot grows suitable area requirement, and energy consumption is reduced to a certain extent.
The present invention is towards the research of preferably Technique for Controlling Greenhouse Environment, in the feasible of explanation multiple objective gray PSO algorithms
During property, selection standard PSO algorithms, two kinds of conventional best practices of linear weight sum method are compared, respectively to temperature, humidity, consumption
Electricity optimizes emulation, draws the comparison form of temperature, humidity and energy consumption, as shown in table 1：
The results contrast of table 1
Table 1 Comparison of results
As shown in Table 1, the temperature and humidity value obtained with standard PSO and multiple objective gray PSO algorithm optimization methods is equal
In suitable temperature humidity range, the temperature and humidity value that linear weight sum method optimization method is obtained suitable temperature humidity range it
Outside.Specific consumption electricity is understood, added with the electricity of multiple objective gray PSO algorithm optimization post consumptions between standard PSO with linear
Between power and method.Preferably strategy is reached simultaneously according to multiple target, and greenhouse is optimized from multiple objective gray PSO algorithms,
In the case where opening combined situation of the sunshade with spray, obtain making temperature and humidity in greenhouse to meet crop optimal growth conditions and power consumption
The less adjusting device combination variety of amount.
Claims (1)
1. the greenhouse optimal control method based on multiple objective gray particle cluster algorithm, it is characterised in that：It is artificial by introducing
Governing factor, based on the autoregression model ARX of extension, builds temperature, humidity and energy consumption cost multi-objective Model function,
On the basis of this, using grey correlation theory and particle swarm optimization algorithm PSO, multiple-objection optimization control is carried out towards greenhouse model
System, specifically includes following steps：
(1), the temperature of multipoint acquisition greenhouse, humidity information, with temperature of the adaptive weighted Fusion Estimation Algorithm to collection
Degree, humidity information carry out data fusion in data Layer, for greenhouse modeling and optimal control；
(2), by introducing artificial governing factor, based on the autoregression model ARX of extension, distinguished with the method for System Discrimination
Know the structure and parameter for model, build chamber environment temperature, the autoregression model of the extension of humidity, and with greenhouse adjusting device
The electricity consumed in operation is that energy consumption cost model is set up in reference, and detailed process is as follows：
(2.1) chamber environment temperature, the autoregression model of the extension of humidity are set up：
The temperature of greenhouse external environment, humidity, wind speed and intensity of illumination are considered as disturbance input amount, by greenhouse adjusting device blower fan,
Wet curtain, sunshade, spray, roof window are considered as decision-making input quantity, and output variable is the temperature and humidity in greenhouse, and greenhouse ring is set up with this
The autoregression model structure of the extension of border epidemic disaster, the form such as following equation institute of the autoregression model of the extension of epidemic disaster
Show：
In formula (1)-(3), y=[y1 y2]TIt is the temperature and humidity in greenhouse；X=[x1 x2…..x9]TFor mode input
Amount；x1For outdoor temperature, unit for DEG C；x2For outside relative humidity；x3For outdoor wind speed, unit is ms-1；x4It is strong for illumination
Degree, unit is lux；x5To x9Respectively greenhouse adjusting device blower fan, wet curtain, sunshade, spray, roof window；A(z-1)、B(z-1) be
ARX model coefficient polynomial, wherein A (z-1) it is the corresponding parametric polynomial of output variable, Bi(z-1) it is respectively nine input changes
Measure corresponding parametric polynomial[12]；z-1For backward shift operator；K is time variable, and unit is min；V (k) is random noise；naFor A
(z-1) rank；nb1,nb2.....nb9For B1(z-1)、B2(z-1).....B9(z-1) rank[12]；Because green-house scale is small, it is considered to v (k)
For significantly or be difficult to guess, then formula (1) can be reduced to：
A(z-1)*yi(t)=B (z-1)*xj(t) (i=1,2；J=1,2,3.....9) (4),
According to actual greenhouse situation, highest order is arranged to second order, and bring formula (2), (3) into formula (4), you can release
The expression of greenhouse epidemic disaster is：
Known z-1* u (t)=u (t-1), z-2* u (t)=u (t-2), substituted into (5) and (6) formula can try to achieve greenhouse temperature,
The autoregression model function expression of the extension of humidity；
(2.2) greenhouse energy consumption cost model is set up：
In green house control equipment running process, the electricity of adjusting device consumption is made up of two parts, and a part is continuously controllable sets
The electricity that received shipment row is consumed for a period of time, another part is the electricity that discontinuous controllable device fully opens consumption, energy consumption cost
Model J can be expressed as：
In formula (7),The electricity of consumption is opened for the regulating and controlling mechanism of common n1 discontinuous actions；For common n2
The electricity of individual continuous action regulating and controlling mechanism consumption；piFor the rated power of single continuous regulating and controlling mechanism；qiFor single discontinuous regulation and control
The rated power of mechanism；xiFor the on off state of regulating and controlling mechanism；taiFor the full opening of run time of noncontinuity regulating and controlling mechanism；
tbiFor continuity regulating and controlling mechanism run time；
(3), by introducing grey correlation theory, on the basis of standard grey relevance theory and particle swarm optimization algorithm PSO,
Adjusting device combination variety is considered as to the solution of particle, it is complete with this using temperature model, Humidity Model and energy consumption model as object function
Into the multiobjective optimal control of greenhouse flower, detailed process is as follows：
(3.1), temperature, humidity and financial cost target are obtained respectively with grey correlation theory and particle swarm optimization algorithm PSO
Adaptive optimal control value of the function under the constraint of control device combination variety, the sequence being made up of the optimal value of each object function is as base
Quasi- vector sequence；
(3.2), according to greenhouse actual conditions, using five kinds of adjusting device permutation and combination species as the solution of particle, in combination variety
Constraints under, initialize each particle scale, iterations, position and speed etc., the initial value of each particle solution converted
Each object function is substituted into binary matrix form, initial target value is obtained, regard this desired value as target vector sequence；
(3.3) degree of association between the base vector sequence and target vector sequence of each particle, is evaluated, by current goal vector
The individual extreme value that sequence degree of association corresponding with its is set to particle is stored as pbest, by the position of the maximum particle of the degree of association and
Its degree of association is stored as gbest as the global extremum of whole population；
(3.4) particle position, speed, Studying factors, weights, are updated, when updating particle position, particle position is rounded
Operation；
(3.5), calculate the degree of association of each particle, by its degree of association with after desired positions corresponding to degree of association work compare
Compared with if more excellent, current degree of association particle position corresponding with its is stored as pbest；Compare the individual extreme value of all particles
And global extremum, if more excellent, the individual extreme value is updated to global extremum, and store its corresponding particle position；
(3.6), judge whether current iteration number of times meets stop condition, if 4. be unsatisfactory for return to step continues iteration；Otherwise, stop
Only search for, the corresponding switchgear combination of output global extremum and temperature, humidity and financial cost value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710173446.8A CN107037728A (en) | 2017-03-22 | 2017-03-22 | Greenhouse optimal control method based on multiple objective gray particle cluster algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710173446.8A CN107037728A (en) | 2017-03-22 | 2017-03-22 | Greenhouse optimal control method based on multiple objective gray particle cluster algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107037728A true CN107037728A (en) | 2017-08-11 |
Family
ID=59534616
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710173446.8A Pending CN107037728A (en) | 2017-03-22 | 2017-03-22 | Greenhouse optimal control method based on multiple objective gray particle cluster algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107037728A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107809461A (en) * | 2017-09-21 | 2018-03-16 | 中国农业大学 | It is a kind of based on high in the clouds towards the management-control method of greenhouse cluster, system and server |
| CN110333171A (en) * | 2019-08-12 | 2019-10-15 | 山东科技大学 | Coal handling system transfer point dust concentration detecting method and system |
| CN112524751A (en) * | 2020-12-01 | 2021-03-19 | 西安建筑科技大学 | Dynamic air conditioning system energy consumption prediction model construction and prediction method and device |
| CN113962075A (en) * | 2021-10-14 | 2022-01-21 | 华翔翔能科技股份有限公司 | Energy management and control method for agricultural planting greenhouse system |
| CN114879769A (en) * | 2022-05-31 | 2022-08-09 | 友联船厂（蛇口）有限公司 | Cabin humidity control method and device |
| CN116384979A (en) * | 2023-04-27 | 2023-07-04 | 圣麦克思智能科技(江苏)有限公司 | IDC operation and maintenance service support system and method thereof |
| CN117289611A (en) * | 2023-11-24 | 2023-12-26 | 中铁建大桥工程局集团第三工程有限公司 | Air conditioning method and system based on fuzzy theory |
| CN118466223A (en) * | 2024-07-12 | 2024-08-09 | 山东科技大学 | Integrated optimization and energy-saving prediction method for cutter and technological parameters in side milling |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104359192A (en) * | 2014-11-19 | 2015-02-18 | 山东建筑大学 | Data-based personalized control system and method for energy conservation and comfort of indoor environment |
| CN106371318A (en) * | 2016-10-31 | 2017-02-01 | 南京农业大学 | Facility environment multi-objective optimization control method based on cooperative game |
-
2017
- 2017-03-22 CN CN201710173446.8A patent/CN107037728A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104359192A (en) * | 2014-11-19 | 2015-02-18 | 山东建筑大学 | Data-based personalized control system and method for energy conservation and comfort of indoor environment |
| CN106371318A (en) * | 2016-10-31 | 2017-02-01 | 南京农业大学 | Facility environment multi-objective optimization control method based on cooperative game |
Non-Patent Citations (6)
| Title |
|---|
| 伍德林 等: ""基于经济最优目标的温室环境控制策略"", 《农业机械学报》 * |
| 唐亚鹏: ""基于自适应加权数据融合算法的数据处理"", 《计算机技术与发展》 * |
| 张武 等: ""基于ARX模型的温室温度模拟"", 《江苏农业学报》 * |
| 张武 等: ""基于灰色粒子群算法的发动机悬置系统多目标优化设计"", 《机械设计》 * |
| 张阳 等: ""基于多传感器自适应加权融合的温室信息系统"", 《传感器与微系统》 * |
| 张雪花 等: ""农业温室环境控制方法研究综述"", 《控制工程》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107809461A (en) * | 2017-09-21 | 2018-03-16 | 中国农业大学 | It is a kind of based on high in the clouds towards the management-control method of greenhouse cluster, system and server |
| CN110333171A (en) * | 2019-08-12 | 2019-10-15 | 山东科技大学 | Coal handling system transfer point dust concentration detecting method and system |
| CN112524751A (en) * | 2020-12-01 | 2021-03-19 | 西安建筑科技大学 | Dynamic air conditioning system energy consumption prediction model construction and prediction method and device |
| CN112524751B (en) * | 2020-12-01 | 2022-04-19 | 西安建筑科技大学 | Dynamic air conditioning system energy consumption prediction model construction and prediction method and device |
| CN113962075A (en) * | 2021-10-14 | 2022-01-21 | 华翔翔能科技股份有限公司 | Energy management and control method for agricultural planting greenhouse system |
| CN114879769A (en) * | 2022-05-31 | 2022-08-09 | 友联船厂（蛇口）有限公司 | Cabin humidity control method and device |
| CN114879769B (en) * | 2022-05-31 | 2023-03-14 | 友联船厂（蛇口）有限公司 | Cabin humidity control method and device |
| CN116384979A (en) * | 2023-04-27 | 2023-07-04 | 圣麦克思智能科技(江苏)有限公司 | IDC operation and maintenance service support system and method thereof |
| CN117289611A (en) * | 2023-11-24 | 2023-12-26 | 中铁建大桥工程局集团第三工程有限公司 | Air conditioning method and system based on fuzzy theory |
| CN118466223A (en) * | 2024-07-12 | 2024-08-09 | 山东科技大学 | Integrated optimization and energy-saving prediction method for cutter and technological parameters in side milling |
| CN118466223B (en) * | 2024-07-12 | 2024-09-17 | 山东科技大学 | Integrated optimization and energy-saving prediction method for cutter and technological parameters in side milling |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107037728A (en) | Greenhouse optimal control method based on multiple objective gray particle cluster algorithm | |
| CN106842923B (en) | Greenhouse environment multi-factor coordination control method based on crop physiology and energy consumption optimization | |
| CN107219759B (en) | Greenhouse environment control method and device | |
| CN101315544A (en) | Greenhouse intelligent control method | |
| CN108983849A (en) | It is a kind of to utilize compound extreme learning machine ANN Control greenhouse method | |
| CN110463587A (en) | Scale greenhouse water-saving irrigation system based on LoRa and GA-BP | |
| CN113554522A (en) | Vineyard accurate drip irrigation control system based on dynamic neural network | |
| Bandara et al. | Smart irrigation controlling system for green roofs based on predicted evapotranspiration | |
| CN112527037A (en) | Greenhouse environment regulation and control method and system with environment factor prediction function | |
| CN112947648B (en) | Agricultural greenhouse environment prediction method and system | |
| CN115633622A (en) | Intelligent orchard irrigation system and method | |
| CN115796040A (en) | Facility greenhouse temperature prediction method based on small samples | |
| CN108319134A (en) | A kind of greenhouse environment intelligent control method based on extreme learning machine network | |
| CN104216445A (en) | Greenhouse flower auto-irrigation control system and control method thereof | |
| Lyu et al. | Design of irrigation control system for vineyard based on lora wireless communication and dynamic neural network | |
| CN116449897B (en) | Greenhouse environment optimal control method, server and system | |
| Li et al. | Temperature prediction model for solar greenhouse based on improved BP neural network | |
| Contreras et al. | Automatic Lighting Control and IoT Monitoring on an Indoor-Greenhouse | |
| CN108415256A (en) | A kind of Cultural control system towards even Donges luminous energy chamber crop | |
| Oktarina et al. | Solar Powered Greenhouse for Smart Agriculture | |
| Zhao et al. | Learning greenhouse climate control policy from monitored data | |
| Simon | Internet of Things and Fuzzy Logic based Smart Balcony Plant Care System from a Renewable Resources | |
| CN108073204A (en) | Agricultural environment control unit | |
| Chen et al. | Optimizing Semi-Closed Greenhouse Temperature, Humidity, CO 2 Concentration, and Light Intensity Using a Nonlinear Model Predictive Control Approach | |
| CN107015480A (en) | A kind of intelligent greenhouse irrigation system based on generalized predictive control and Internet of Things |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170811 |