CN103558888A - Intelligent control system and method for environment in cowshed - Google Patents

Abstract

The invention provides an intelligent control system and method for the environment in a cowshed. A temperature and humidity sensor inside the cowshed and a temperature and humidity sensor outside the cowshed, a gas analyzer and a wind speed gauge are taken as signal sources of the intelligent control system, and a circulating fan inside the cowshed, a roll-up window driving motor and a spraying driving motor are taken as executing equipment of the intelligent control system. Intelligent decision and reasoning of the environment in the cowshed can be achieved according to status of environment variables by control equipment with the case-based reasoning and group decision intelligent reasoning method; group decision correction can be carried out on a decision result of a target case; start and stop commands of the equipment are automatically output, and the automatic control function is achieved. The control system is simple in hardware structure, fast in response, high in measurement accuracy and free from being supervised by specially-assigned persons. The environment in the cowshed is controlled with the control method and the case-based reasoning and group decision intelligent algorithm, and thus the control system has very high self-adaptation and self-learning capacity, the cowshed can be controlled in a relatively stable and comfortable environment, and heavy losses caused by milk yield reduction and even sickness and death resulting from deterioration of an environment where cows are located are avoided.

Description

Cowshed Intelligent environment control system and control method

Technical field

The present invention relates to automation field, particularly a kind of cowshed Intelligent environment control system and control method.

Background technology

Milk contains abundant mineral matter, calcium, and the content of phosphorus, iron, zinc, copper, manganese, molybdenum is all a lot.The rareest, milk is human calcium's best source, and calcium phosphorus ration is very suitable, is beneficial to the absorption of calcium.Along with the more and more scale of aquaculture of milk cow, the continuous quickening of modernization, also more and more higher to the requirement of the each side such as the scale of cowshed, quality, environment control degree.Experimental result shows: in house, temperature can be controlled in-10 ℃～27 ℃, and indoor humidity can be controlled in 50～70%RH, and the content of harmful gas can be controlled in CO ₂<0.15%, H ₂s<15mg/m3, NH ₃<19.5mg/m3, the output of milk and quality can reach higher level.Therefore, the application of cowshed environment controlling technique has important practical significance.

Because cowshed volume is large, get rid of harmful gas and cool-down dehumidification device distribution numerous, operational decisions difficulty to these equipment is very large, the level that is difficult to environment in house to be adjusted in real time by operating personnel or conventional control device a relative comfort, thereby, need the consideration import of advanced technology and method to go to capture this difficult problem.Since the eighties in 20th century, a kind of new problem solving method-reasoning by cases (Case-Based Reasoning in artificial intelligence, CBR) obtained extensive attention and research, it is adapted at being difficult to set up mechanism model and domain knowledge not exclusively and is difficult to definition but uses in veteran policy setting.CBR is as a kind of new problem solving and machine learning method, reasoning process can be described with classical 4R cognitive model, i.e. the retrieval of case (Retrieve), reuse (Reuse), revise (Revise) and store (Retain).Wherein, the correction stage plays vital effect to the quality of problem solving, directly affects the correctness of the reasoning results, is a still unsolved open problem.Therefore,, when application CBR method, need to consider the method for case correction, to obtain the reasoning results accurately.

Summary of the invention

The technical matters that the present invention mainly solves is, a kind of intelligent control method that can regulate in real time environment in cowshed is provided, can be according to the state variation of humiture, harmful gas content and outdoor temperature humidity, wind speed, use reasoning by cases and group decision intelligent algorithm, to giving up the start and stop of the equipment such as interior circulating fan, roll-up window, spray, control, and then provide life, a growing environment that health is suitable for milk cow.The present invention's the first object is to propose a kind of cowshed ambient intelligence control method, it is characterized in that, said method comprising the steps of:

Step 1, initialization of variable;

Step 2, set up case library; Source case C in case library _k=(X _k; Y _k), k=1,2 ..., p, wherein, p is source case sum; X _k=(x _{1, k}, x _{2, k}..., x _{8, k}) and Y _k=(y _{1, k}, y _{2, k}..., y _{5, k}) be respectively problem description and the case answer of k source case; x _i,k(i=1,2 ..., 8) and represent the state value of i the cowshed environmental characteristic variable that problem in k source case is described; y _l,k(l=1,2 ..., 5) and represent the start and stop state of l cowshed actuating equipment in the case answer of k source case;

Step 3, judge whether to occur new target case, if NO, jump procedure 10; Otherwise, obtain the problem of target case and describe: X _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}), wherein, x _{i, p+1}(i=1,2 ..., 8) and represent that the problem of target case describes X _p+1in the state value of i cowshed environmental characteristic variable, and perform step 4;

Step 4, source case problem is described to X _k=(x _{1, k}, x _{2, k}..., x _{8, k}), k=1,2 ..., p and target case problem are described X _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}) in the state value of cowshed environmental characteristic variable be normalized;

Step 5, according to target case problem describe describe with source case problem in the state value of each cowshed environmental characteristic variable after normalization, calculate the similarity s of each source case and target case _k;

Step 6, according to the similarity threshold s setting _v, determine the number of coupling case, if jump procedure 7 while mating case number m>1; Otherwise, select described similarity s _kmaximal value the case answer of corresponding source case as the case answer Y of target case _p+1, wherein, Y _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}), y _{l, p+1}(l=1,2 ..., 5) represent the start and stop state of l cowshed actuating equipment, and go to step 8.

The start and stop state of step 7, group decision corrective, obtains the case of this target case and answers Y _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1});

Step 8, case are reused;

Step 9, by goal-trail leading case example answer Y _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}) and normalization before target case problem X is described _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}) form new source case C _p+1=(X _p+1; Y _p+1) be stored in case library, and source case sum p is increased to 1;

If step 10 control task finishes, EOP (end of program); Otherwise, go to step 3.

Preferably, described cowshed environmental characteristic variable is outdoor temperature, outside humidity, outdoor wind speed, indoor temperature, indoor humidity, indoor CO ₂, NH ₃, H ₂s concentration.

Preferably, described cowshed actuating equipment comprises circulating fan in house, upper roll-up window drive motor, lower roll-up window drive motor, south spray drive motor, north spray drive motor.

Preferably, the value of the start and stop state of cowshed actuating equipment is 0 or 1.

Preferably, normalized described in step 4 is undertaken by following formula:

${\tilde{x}}_{i,k}=\frac{x_{i,k}-\min (x_{i,1},...,x_{i,p+1})}{\max (x_{i,1},...,x_{i,p+1})-\min (x_{i,1},...,x_{i,p+1})},i=1,...,8;k=\mathrm{1,2},...,p$

${\tilde{x}}_{i,p+1}=\frac{x_{i,p+1}-\min (x_{i,1},...,x_{i,p+1})}{\max (x_{i,1},...,x_{i,p+1})-\min (x_{i,1},...,x_{i,p+1})},i=1,...,8$

Wherein,

state value for i cowshed environmental characteristic variable in the problem description of k source case after normalization; state value for i cowshed environmental characteristic variable in the problem description of target case after normalization.

Preferably, in step 5, be calculated as follows the similarity s of each source case and target case _k:

$s_k=1-\sqrt{\frac{1}{8}\underset{i=1}{\overset{8}{\mathrm{\&Sigma;}}}{({\tilde{x}}_{i,p+1}-{\tilde{x}}_{i,k})}^2},k=\mathrm{1,2},...,p$

Preferably, in step 6, s _v∈ (0,1], similarity is more than or equal to s _vsource case be the coupling case of target case.

Preferably, in step 7, obtain the case answer Y of this target case _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}) according to following formula, obtain:

$y_l=\frac{\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{s}_j{y}_{l,j}}{\underset{j}{\overset{m}{\mathrm{\&Sigma;}}}{s}_j},l=\mathrm{1,2},...,5$

$\left\{\begin{array}{c}{y}_{l,p+1}=1,y_l>0.5\\ y_{l,p+1}=0,y_l<0.5\end{array}\right.,l=\mathrm{1,2}...,5$

S _j(j=1,2 ..., m) represent j similarity size of mating case and target case in m coupling case;

Y _l,j(l=1,2 ..., 5) represent in m coupling case j case of mating case answer in the start and stop state of l cowshed actuating equipment; y _lthe start and stop probability that represents l platform equipment.

Another object of the present invention is to propose a kind of cowshed Intelligent environment control system, comprises CO in She Nei and the outer Temperature Humidity Sensor of house, house ₂h in gas analyzer, house ₂nH in S gas analyzer, house ₃gas analyzer, air speed measuring apparatus and PLC controller, analog input module, digital quantity input module, digital output module, power module, touch-screen, monitor terminal and relay module, it is characterized in that indoor temperature and humidity sensor, outdoor temperature humidity sensor, air speed measuring apparatus, CO ₂gas analyzer, H ₂s gas analyzer, NH ₃gas analyzer is connected with analog input module respectively, obtains indoor temperature, indoor humidity, outdoor temperature, outside humidity, outdoor wind speed and indoor CO ₂concentration, H ₂s concentration, NH ₃concentration; Digital output module is connected with cowshed actuating equipment by relay, and controls its start and stop; Each actuating equipment is connected with relay simultaneously, sends run signal to digital quantity input module; PLC controller is connected with touch-screen by RS232 serial port respectively, by Ethernet interface, is connected with monitor terminal.

Preferably, described cowshed actuating equipment comprises circulating fan in house, upper roll-up window drive motor, lower roll-up window drive motor, south spray drive motor and north spray drive motor.

Intelligent control system of the present invention is by the interior CO of Temperature Humidity Sensor, house that is distributed in a plurality of check points in house ₂h in gas analyzer, house ₂nH in S gas analyzer, house ₃the Temperature Humidity Sensor of gas analyzer, the outer check point of house, air speed measuring apparatus are as signal source, by programmable logic controller (PLC) (Programmable Logic Controller, PLC), analog input module, digital quantity input module, digital output module, RS232 serial port, Ethernet interface, monitor terminal, touch-screen, power module, relay etc. are as opertaing device, by giving up interior circulating fan, upper (lower) roll-up window drive motor and south (north) side spray pouring drive motor as actuating equipment.Its control method adopts reasoning by cases and group decision intelligent inference method, on the basis of traditional 4R cognitive model, uses group decision thought to go to revise case answer.Can realize according to the real-time status of environmental variance intelligent decision and the reasoning of environment in house; To the result of decision of target case, can realize group decision correction; Automatically the start stop command of output device, realizes automatic control function.

The system architecture of the technical program is simple, fast response time, a plurality of environmental parameters and the measuring accuracy that can measure in cowshed are high, has very strong self-adaptation and self-learning capability, need not special messenger keep an eye on and operate, and saves labour cost, has significantly reduced cost; Can make cowshed be controlled at a relatively stable and comfortable environment; Can avoid milk cow growth, living environment to worsen and cause its output of milk to reduce even sick death and the heavy losses that cause.

Accompanying drawing explanation

Fig. 1 is the structural representation of cowshed Intelligent environment control system of the present invention.

Fig. 2 is the case based reasoning model structure principle chart based on group decision correction.

Fig. 3 is the schematic flow sheet of cowshed ambient intelligence control method of the present invention.

Embodiment

Below in conjunction with drawings and Examples, the invention will be further described.

With reference to Fig. 1, cowshed Intelligent environment control system and implementation method, by the Temperature Humidity Sensor (1), the CO that are distributed in house ₂gas analyzer (3), H ₂s gas analyzer (4), NH ₃gas analyzer (5), be distributed in outdoor Temperature Humidity Sensor (2), air speed measuring apparatus (6) is connected with analog input module (7) as signal source, obtains indoor temperature, indoor humidity, indoor CO ₂, NH ₃, H ₂the measured value of 8 characteristic variables such as S concentration, outdoor temperature, outside humidity, outdoor wind speed, digital output module (9) is by the interior circulating fan (10) of relay (22) and house, upper roll-up window drive motor (11), lower roll-up window drive motor (12) and south spray drive motor (13), the cowshed actuating equipments such as north spray drive motor (14) are connected, export the start and stop state of these cowshed actuating equipments, the start and stop status signal of above-mentioned each cowshed actuating equipment sends digital quantity input module (8) to by relay (21), PLC controller (15) is connected with touch-screen (19) by RS232 serial port (18) respectively, by Ethernet interface (16), be connected with monitor terminal (17), power module (20) is given touch-screen, PLC, monitor terminal provides power supply.Control software in PLC controller is realized monitoring and the adjustment of equipment state, variable parameter.

Consulting Fig. 2, is the case based reasoning model structure principle chart based on group decision correction.On the basis of traditional 4R cognitive model, use Group Decision Theory to revise case answer.

In case library, store be respectively organized the characteristic variable value of cowshed environmental parameter in the past and the in the situation that of this stack features variate-value the example of relevant device start and stop state, be called source case, as cowshed environmental characteristic variable x _{i, p+1}(i=1,2 ..., 8) and the target case problem that forms describes X _p+1input to after system, need to infer the case answer Y of this target case _p+1, i.e. the start and stop status command set of 5 kinds of cowshed actuating equipments shown in Fig. 1, wherein Y _p+1by y _{l, p+1}(j=1,2, ... 5) form, be respectively circulating fan (10) in house, upper roll-up window drive motor (11), lower roll-up window drive motor (12), south spray drive motor (13), north spray drive motor (14) start and stop state separately, 1 represents to start, and 0 represents to stop.By Case Retrieval, calculate the problem of target case and describe X _p+1with every source case problem in case library, X is described _k(k=1,2 ... similarity s p) _k, total p, and set a similarity threshold s _vwith the p a calculating s _kcompare, meet s _k>=s _vtime corresponding source case be called coupling case, now have two kinds of results and occur: the one, coupling case number has when a plurality of, reuses after adopting group decision thought to revise case answer; The 2nd, only have one or while not mating case, need only select p s _kin source case corresponding to maximum similarity case answer duplicate removal with.Be finally case storage, this target case problem is described to X _p+1and corresponding case answer Y _p+1form new source case C _k+1, be stored in case library, for reasoning, solve the answer of next target case.

Consulting Fig. 3, is the schematic flow sheet of cowshed Intelligent environment control system of the present invention and its implementation.As can be seen from the figure, concrete implementation step is as described below:

Step 1, initialization of variable.Set similarity threshold s _v, s _v∈ (0,1].

Step 2, set up case library.Source case representation in storehouse is C _k=(X _k; Y _k), k=1,2 ..., p, wherein, p is source case sum; X _k=(x _{1, k}, x _{2, k}..., x _{8, k}) and Y _k=(y _{1, k}, y _{2, k}..., y _{5, k}) be respectively problem description and the case answer of k source case; x _i,k(i=1,2 ..., 8) and represent the state value of i the characteristic variable that problem in k source case is described, distinguish corresponding outdoor temperature, outside humidity, outdoor wind speed, indoor temperature, indoor humidity, indoor CO ₂, NH ₃, H ₂s concentration; y _l,k(l=1,2,5) represent the start and stop state of l platform equipment in the case answer of k source case, correspondence is given up interior circulating fan, upper roll-up window drive motor, lower roll-up window drive motor, south spray drive motor, the northern start and stop state that sprays drive motor respectively, 1 represents to start, and 0 represents to stop.

Step 3, judge whether to occur new target case, if nothing goes to step 10; Otherwise, obtain the problem of target case and describe: X _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}), wherein, x _{i, p+1}(i=1,2 ..., 8) and represent that the problem of target case describes X _p+1in the state value of i characteristic variable; Need to obtain the case answer Y of this target case _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}); Wherein, y _{l, p+1}(l=1,2 ..., 5) represent the case answer Y of target case _p+1in the start and stop state of l platform equipment.

Step 4, characteristic variable normalization.Source case problem is described to X _k=(x _{1, k}, x _{2, k}..., x _{8, k}), k=1,2 ..., p and target case problem are described X _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}) in characteristic variable value be normalized, by following formula, undertaken:

${\tilde{x}}_{i,k}=\frac{x_{i,k}-\min (x_{i,1},...,x_{i,p+1})}{\max (x_{i,1},...,x_{i,p+1})-\min (x_{i,1},...,x_{i,p+1})},i=1,...,8;k=\mathrm{1,2},...,p$

${\tilde{x}}_{i,p+1}=\frac{x_{i,p+1}-\min (x_{i,1},...,x_{i,p+1})}{\max (x_{i,1},...,x_{i,p+1})-\min (x_{i,1},...,x_{i,p+1})},i=1,...,8$

Wherein, x _i,k(k=1,2 ..., p) be the actual measurement state value of i characteristic variable in the problem description of front k the source case of normalization, state value for i characteristic variable in the problem description of k source case after normalization; x _{i, p+1}for the actual measurement state value of i characteristic variable in the problem description of target case before normalization,

state value for i characteristic variable in the problem description of target case after normalization.

Step 5, according to the problem of target case describe describe with the problem of source case in the state value of each characteristic variable after normalization, calculate the similarity of each source case and target case, as shown in the formula carrying out:

$s_k=1-\sqrt{\frac{1}{8}\underset{i=1}{\overset{8}{\mathrm{\&Sigma;}}}{({\tilde{x}}_{i,p+1}-{\tilde{x}}_{i,k})}^2},k=\mathrm{1,2},...,p$

Step 6, according to the similarity threshold s setting _v, determine the number of mating case.According to the similarity threshold s setting _v, compare respectively p s _kwith s _vsize, suppose to be more than or equal to s _vcoupling case number be m, during m>1, go to step 7; Otherwise, the similarity s of selection source case and target case _kthe case answer of the corresponding source of maximal value case is answered as the case of target case, and goes to step 8.

The start and stop state of step 7, group decision corrective.If m similarity size of mating case and target case is respectively s _j(j=1,2 ..., m), j case answer corresponding to coupling case is Y _j=(y _{1, j}, y _{2, j}..., y _{5, j}), j=1,2 ..., m, y _l,j(l=1,2 ..., 5) and represent the start and stop state of l platform equipment in j coupling case case answer; Adopt group decision thought to revise the equipment start-stop state of goal-trail leading case example answer, obtain the start and stop probable value of each equipment, y _lthe routine start and stop probability of answering corresponding l platform equipment of goal-trail leading case after representing to revise:

$y_l=\frac{\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{s}_j{y}_{l,j}}{\underset{j}{\overset{m}{\mathrm{\&Sigma;}}}{s}_j},l=\mathrm{1,2},...,5$

If y _l>0.5 (l=1,2 ..., 5), the start and stop state that this equipment is set is 1; Otherwise, set to 0, thereby obtain the case answer Y of this target case _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}), y _{l, p+1}(l=1,2 ..., 5) represent the start and stop state of l cowshed actuating equipment; That is:

$\left\{\begin{array}{c}{y}_{l,p+1}=1,y_l>0.5\\ y_{l,p+1}=0,y_l<0.5\end{array}\right.,l=\mathrm{1,2}...,5$

Step 8, case are reused.By the similarity s obtaining in step 6 _kmaximum source case is carried out direct case and is reused, or according to step 7, adopt group decision thought to revise the equipment start-stop state of goal-trail leading case example answer based on coupling case, being equivalent to that coupling case is carried out to indirectly case reuses, thereby obtain each equipment start-stop state, and the digital output module in Fig. 1 export this order, control corresponding equipment.

Step 9, storage new case.The goal-trail leading case example answer Y that this reasoning is obtained _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}) and normalization before target case problem X is described _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}) form new source case C _p+1=(X _p+1; Y _p+1) be stored in case library, source case sum p increases 1, i.e. p ← p+1.

If step 10 control task finishes, for example, system is cut off, is ended or overhauls, EOP (end of program); Otherwise, go to step 3.

In embodiments of the invention, summed up the source case of controlling for cowshed ambient intelligence, totally 150 records, source case sum p=150, is stored in case library.For a concrete target case, carrying out reasoning below solves.

In step 1, set similarity threshold s _v=0.85;

In step 3, the eigenwert of target case is described as X _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1})=(35.6,68.4,3,33.7,81.2,0.22,13.9,17.3);

In step 4, in order to eliminate the impact of the different dimensions of measurement data, use following formula normalization

${\tilde{x}}_{i,k}=\frac{x_{i,k}-\min (x_{i,1},...,x_{i,p+1})}{\max (x_{i,1},...,x_{i,p+1})-\min (x_{i,1},...,x_{i,p+1})},i=1,...,8;k=\mathrm{1,2},...,p$

${\tilde{x}}_{i,p+1}=\frac{x_{i,p+1}-\min (x_{i,1},...,x_{i,p+1})}{\max (x_{i,1},...,x_{i,p+1})-\min (x_{i,1},...,x_{i,p+1})},i=1,...,8$

After normalization, the eigenwert of target case is described as ${\tilde{X}}_{p+1}=({\tilde{x}}_{1,p+1},{\tilde{x}}_{2,p+1},...,{\tilde{x}}_{8,p+1})=(\mathrm{0.8233,0.7785,0.6500},$ $\mathrm{0.8546,0.8914,0.6539,0.5618,0.4125});$

In step 5, calculate the similarity of target case and source case, as shown in the formula carrying out:

$s_k=1-\sqrt{\frac{1}{8}\underset{i=1}{\overset{8}{\mathrm{\&Sigma;}}}{({\tilde{x}}_{i,p+1}-{\tilde{x}}_{i,k})}^2},k=\mathrm{1,2},...,p$

In step 6, according to the similarity threshold s setting _v=0.85, compare respectively p s _kwith s _vsize, be more than or equal to s _vcoupling case number m=3, similarity size is respectively s ₁=0.93, s ₂=0.89, s ₃=0.88, corresponding case answer Y ₁=(1,1,1,0,0), Y ₂=(1,1,1,1,0), Y ₃=(1,1,0,0,0);

In step 7, according to following formula, the start and stop state of equipment is carried out to group decision correction:

$y_l=\frac{\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{s}_j{y}_{l,j}}{\underset{j}{\overset{m}{\mathrm{\&Sigma;}}}{s}_j},l=\mathrm{1,2},...,5$

Therefrom can obtain the start stop command of each equipment, respectively: circulating fan y in house ₁=1, upper roll-up window drive motor y ₂=1, lower roll-up window drive motor y ₃=0.67, south spray drive motor y ₄=0.33, north spray drive motor y ₅=0.According to the equipment start-stop condition in this step, if y _l>0.5 (l=1,2 ..., 5), the start and stop state that this equipment is set is 1; Otherwise, set to 0.Therefore, the reasoning results is the reasoning results Y _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1})=(1,1,1,0,0), therefrom the start and stop result of equipment is: start circulating fan, upper roll-up window and lower roll-up window drive motor in house, stop the drive motor of south, north side spray;

In step 8, the answer using the equipment start-stop order obtaining as target case, and the digital output module in Fig. 1 is exported operation that this order removes to control cowshed environment actuating equipment and is stopped;

In step 9, by target case X _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1})=(35.6,68.4,3,33.7,81.2,0.22,13.9,17.3) and corresponding the reasoning results Y _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1})=(1,1,1,0,0) be stored in case library, now, the sum of source case becomes 151, carries out reasoning next time and solve while supplying to obtain new target case.

Cowshed Intelligent environment control system and implementation method, it is characterized in that the method is according to the cowshed environment temperature detecting, humidity, harmful gas content, case based reasoning model and the algorithm of employing based on group decision correction, obtain starting the order of field apparatus, by digital output module, start corresponding actuating equipment, the operating state signal of these equipment also can be sent to by digital quantity input module the handling procedure of PLC, all signals, the variable condition of parameter through Ethernet and RS232 serial port respectively at monitor terminal with touch screen display, for operating personnel's observation and analysis.In addition, also can at monitor terminal or touch-screen, send instruction by operating personnel, by digital output module, artificially start the actuating equipment that cowshed environment is controlled, thereby realize man-machine coordination work.

Must be pointed out, above-mentioned example is only made an indefiniteness to the present invention and is illustrated.But person of skill in the art will appreciate that, do not departing under aim of the present invention and scope, can make various modifications, replacement and change to the present invention, these are revised, replace and change still belongs to protection scope of the present invention.

Claims (10)

1. a cowshed ambient intelligence control method, is characterized in that, it is characterized in that, said method comprising the steps of:

Step 1, initialization of variable;

Step 2, set up case library; Source case C in case library _k=(X _k; Y _k), k=1,2 ..., p, wherein, p is source case sum; X _k=(x _{1, k}, x _{2, k}..., x _{8, k}) and Y _k=(y _{1, k}, y _{2, k}..., y _{5, k}) be respectively problem description and the case answer of k source case; x _i,k(i=1,2 ..., 8) and represent the state value of i the cowshed environmental characteristic variable that problem in k source case is described; y _l,k(l=1,2 ..., 5) and represent the start and stop state of l cowshed actuating equipment in the case answer of k source case;

Step 3, judge whether to occur new target case, if NO, jump procedure 10; Otherwise, obtain the problem of target case and describe: X _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}), wherein, x _{i, p+1}(i=1,2 ..., 8) and represent that the problem of target case describes X _p+1in the state value of i cowshed environmental characteristic variable, and perform step 4;

Step 4, source case problem is described to X _k=(x _{1, k}, x _{2, k}..., x _{8, k}), k=1,2 ..., p and target case problem are described X _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}) in the state value of cowshed environmental characteristic variable be normalized;

Step 5, according to target case problem describe describe with source case problem in the state value of each cowshed environmental characteristic variable after normalization, calculate the similarity s of each source case and target case _k;

Step 6, according to the similarity threshold s setting _v, determine the number of coupling case, if jump procedure 7 while mating case number m>1; Otherwise, select described similarity s _kmaximal value the case answer of corresponding source case as the case answer Y of target case _p+1, wherein, Y _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}), y _{l, p+1}(l=1,2 ..., 5) represent the start and stop state of l cowshed actuating equipment, and go to step 8.

The start and stop state of step 7, group decision corrective, obtains the case of this target case and answers Y _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1});

Step 8, case are reused;

Step 9, by goal-trail leading case example answer Y _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}) and normalization before target case problem X is described _p+1=(x _{1, p+1}, x _{2, p+1}..., x _{8, p+1}) form new source case C _p+1=(X _p+1; Y _p+1) be stored in case library, and source case sum p is increased to 1;

If step 10 control task finishes, EOP (end of program); Otherwise, go to step 3.

2. cowshed ambient intelligence control method as claimed in claim 1, is characterized in that, described cowshed environmental characteristic variable is outdoor temperature, outside humidity, outdoor wind speed, indoor temperature, indoor humidity, indoor CO ₂, NH ₃, H ₂s concentration.

3. cowshed ambient intelligence control method as claimed in claim 1, is characterized in that, described cowshed actuating equipment comprises circulating fan in house, upper roll-up window drive motor, lower roll-up window drive motor, south spray drive motor, north spray drive motor.

4. cowshed ambient intelligence control method as claimed in claim 1, is characterized in that, the value of the start and stop state of cowshed actuating equipment is 0 or 1.

5. cowshed ambient intelligence control method as claimed in claim 1, is characterized in that, normalized described in step 4 is undertaken by following formula:

${\tilde{x}}_{i,k}=\frac{x_{i,k}-\min (x_{i,1},...,x_{i,p+1})}{\max (x_{i,1},...,x_{i,p+1})-\min (x_{i,1},...,x_{i,p+1})},i=1,...,8;k=\mathrm{1,2},...,p$

${\tilde{x}}_{i,p+1}=\frac{x_{i,p+1}-\min (x_{i,1},...,x_{i,p+1})}{\max (x_{i,1},...,x_{i,p+1})-\min (x_{i,1},...,x_{i,p+1})},i=1,...,8$

Wherein,

state value for i cowshed environmental characteristic variable in the problem description of k source case after normalization; state value for i cowshed environmental characteristic variable in the problem description of target case after normalization.

6. cowshed ambient intelligence control method as claimed in claim 1, is characterized in that, is calculated as follows the similarity s of each source case and target case in step 5 _k:

$s_k=1-\sqrt{\frac{1}{8}\underset{i=1}{\overset{8}{\mathrm{\&Sigma;}}}{({\tilde{x}}_{i,p+1}-{\tilde{x}}_{i,k})}^2},k=\mathrm{1,2},...,p$

7. cowshed ambient intelligence control method as claimed in claim 1, is characterized in that, in step 6, and s _v∈ (0,1], similarity is more than or equal to s _vsource case be the coupling case of target case.

8. cowshed ambient intelligence control method as claimed in claim 1, is characterized in that, obtains the case answer Y of this target case in step 7 _p+1=(y _{1, p+1}, y _{2, p+1}..., y _{5, p+1}) according to following formula, obtain:

$y_l=\frac{\underset{j=1}{\overset{m}{\mathrm{\&Sigma;}}}{s}_j{y}_{l,j}}{\underset{j}{\overset{m}{\mathrm{\&Sigma;}}}{s}_j},l=\mathrm{1,2},...,5$

$\left\{\begin{array}{c}{y}_{l,p+1}=1,y_l>0.5\\ y_{l,p+1}=0,y_l<0.5\end{array}\right.,l=\mathrm{1,2}...,5$

S _j(j=1,2 ..., m) represent j similarity size of mating case and target case in m coupling case;

Y _l,j(l=1,2 ..., 5) represent in m coupling case j case of mating case answer in the start and stop state of l cowshed actuating equipment; y _lthe start and stop probability that represents l platform equipment.

9. a cowshed Intelligent environment control system, comprises CO in She Nei and the outer Temperature Humidity Sensor of house, house ₂h in gas analyzer, house ₂nH in S gas analyzer, house ₃gas analyzer, air speed measuring apparatus and PLC controller, analog input module, digital quantity input module, digital output module, power module, touch-screen, monitor terminal and relay module, it is characterized in that indoor temperature and humidity sensor, outdoor temperature humidity sensor, air speed measuring apparatus, CO ₂gas analyzer, H ₂s gas analyzer, NH ₃gas analyzer is connected with analog input module respectively, obtains indoor temperature, indoor humidity, outdoor temperature, outside humidity, outdoor wind speed and indoor CO ₂concentration, H ₂s concentration, NH ₃concentration; Digital output module is connected with cowshed actuating equipment by relay, and controls its start and stop; Each actuating equipment is connected with relay simultaneously, sends run signal to digital quantity input module; PLC controller is connected with touch-screen by RS232 serial port respectively, by Ethernet interface, is connected with monitor terminal.

10. a cowshed Intelligent environment control system as claimed in claim 9, is characterised in that, described cowshed actuating equipment comprises circulating fan in house, upper roll-up window drive motor, lower roll-up window drive motor, south spray drive motor and north spray drive motor.

Images