TW201909093A - Intelligent scheduling system and method thereof comprising a real-time data, a historical database, an analysis module, an automatic watering module, and a parameter feedback database - Google Patents

Abstract

An intelligent scheduling system is suitable for watering a crop, comprising a real-time data, a historical database, an analysis module, an automatic watering module, and a parameter feedback database. The real-time data are a plurality of real-time external environment variable data of the crop, the historical database stores a plurality of historical growth data of the crop, the analysis module stores a plurality of the watering schedule parameter data prepared by processing the data of the real-time external environmental variable data and the historical growth data. The automatic watering module is connected with the analysis module, and includes an automatic watering machine capable of receiving the watering schedule parameter data. The parameter feedback database stores plural parameter feedback data of the automatic watering machine to the crop in the watering process and the parameter adjustment data of manual adjustment.

Description

 Intelligent scheduling system and method thereof  

The present invention relates to a scheduling system, and more particularly to an intelligent scheduling system and method thereof.

Irrigation equipment traditionally used in agriculture or planting is to manually monitor the current soil conditions and irrigate them according to actual needs. In the work site of the Lan Garden, the manpower spray is still the main one, and the time course and water volume of the spray are also based on the assessment of the on-site personnel. If the staff is transferred, different judgments will be made between the new and old personnel. Unnecessary waste of resources due to human factors.

Referring to FIG. 1, an automatic sprinkler system and method thereof is disclosed in the Chinese Patent No. 201519954. The automatic sprinkler system 100 includes a carrier member 102 on which an image for capturing an image signal is respectively disposed. The imaging component 104, an image analyzing component 106 for analyzing the image signal, a nozzle 108, and a first actuating component 110 for adjusting the opening direction of the nozzle 108, the first actuating component 110 is adjusted during operation. The opening direction of the nozzle 108 is such that the opening direction of the nozzle 108 faces the predetermined spraying area recognized by the image analyzing component 106 to accurately complete the spraying action.

From the above description, it is known that the conventional automatic sprinkler system and its method still have the following disadvantages in actual use:

First, the irrigation effect is not good

Crops in the same irrigation area may have different soil moisture requirements due to external environmental factors, such as the angle of sunlight. Therefore, the amount of irrigation water must be controlled according to different humidity requirements to meet the ideal soil moisture environment of crops. The spraying system can only perform image analysis on the surface of the crop. When the temperature is too high or the humidity is insufficient, it is impossible to automatically detect and start the nozzle 108 for watering, which is obviously insufficient to meet the irrigation demand.

Second, lack of functionality

The monitoring of the surrounding environment of cultivated crops and the management of crop cultivation is an extremely important task. It is only necessary to carry out image analysis on the changes of crops themselves, and does not monitor the surrounding environment as a manager. The basis for treatment or spraying is not easy for ordinary farmers to use.

The above disadvantages all show various problems arising from the use of the conventional automatic spraying system and its method, such as improving the existing spraying method, designing to reduce the influence of human factors and resource consumption, and at the same time improving the stability of orchid growth. Sex will enhance the competitiveness of the market.

In view of the above, an object of the present invention is to provide an intelligent scheduling system suitable for watering a crop, comprising an instant data, a historical database, an analysis module, an automatic watering module, and A parameter feedback database.

The real-time data is a plurality of real-time external environmental variables of the crop, the historical database storing a plurality of historical growth data of the crop, the analysis module is connected with the real-time data and the historical database, and stored in plural The watering schedule parameter data prepared by processing the real-time external environment variable data and the historical growth data, the automatic watering module is connected with the analysis module, and includes a data for receiving the watering schedule parameter The automatic watering machine, the parameter feedback database is connected with the analysis module and the automatic watering module, and stores a plurality of parameter feedback data of the watering operation process of the automatic watering machine on the crop.

Another technical means of the present invention is that the intelligent scheduling system further includes a remote control module, which includes a communication interface connected to the analysis module, for transmitting the watering schedule parameter data to the automatic Watering machine.

A further technical means of the present invention is that the automatic watering module further comprises a controller disposed on the automatic watering machine, wherein a user manually adjusts the operation of the automatic watering machine, and the parameter feedback data The library further stores a plurality of parameter adjustment data for the controller to operate.

A further technical means of the present invention is that the above-mentioned immediate external environmental variable data records information such as temperature, humidity, atmospheric pressure, soil, lighting, fan, and cold air of the environment in which the crop is located.

Another technical means of the present invention is that the historical growth data record has a pre-scheduled watering application process for the crop, the crop application information, the crop scrap information, the crop breeding rate, and the history. Immediate external environmental variables and other information.

Another object of the present invention is to provide a method for performing the above-described intelligent scheduling system, the intelligent scheduling method comprising a collecting step, a data processing step, an applying step, and a parameter feedback step.

First, the collecting step is performed, the environmental variable data is collected in the environment outside the crop, and the instant external environment variable data is transmitted to the analysis module, and then the data processing step is performed, and the analysis module is configured to The immediate external environmental variable data of the crop, and the historical growth data are subjected to data processing operations to obtain the watering schedule parameter data, and transmitted to the automatic watering machine, and then the performing step is performed, when the automatic watering machine receives After the scheduling parameter data is poured, the automatic watering machine performs watering and applying the crop according to the watering schedule parameter data, and finally, the parameter feedback step is performed, and after the automatic watering machine waters the crop, The parameter feedback data will be sent and sent to the parameter feedback database.

Another technical means of the present invention is that the data processing step is to transmit the watering schedule parameter data to the automatic watering machine by using the communication interface.

A further technical means of the present invention is the above-mentioned intelligent scheduling method, further comprising an empirical learning step after the parameter feedback step, in which the parameter feedback data is transmitted to the analysis module The data processing operation of the data processing step is performed.

Another technical means of the present invention is the intelligent scheduling method described above, further comprising a manual adjustment step between the data processing step and the applying step, in the manual adjusting step, when the automatic watering machine receives the Before watering the scheduling parameter data, and before the irrigation of the crop has been applied, the user can manually adjust the governor to change the watering parameters of the automatic irrigation machine for watering the crop, and prepare the parameter adjustment data. And then transferring the parameter adjustment data to the data processing operation of the data processing step in the analysis module.

Another technical means of the present invention is that in the above-mentioned data processing steps, it refers to data filtering, data analysis, data normalization, data training, and data collection and the like.

The beneficial effects of the present invention are: using the instant external environment variable data and the historical scheduling data to make the watering schedule parameter data, the parameter feedback information made by the crop after the watering application, and the user manually adjusting the governor The parameter adjustment data is made, and the analysis module is fed back to the machine for learning, and as the empirical learning parameter of the next watering application, a set of standard implementation strategy is established, and the automation intelligence can be performed without relying on the help of the expert. agriculture.

2‧‧‧ Crops

3‧‧‧ Real-time data

31‧‧‧ Immediate external environmental variables

4‧‧‧Historical database

41‧‧‧ Historical growth data

5‧‧‧Analysis module

51‧‧‧Watering schedule parameters

6‧‧‧Automatic watering module

61‧‧‧Automatic watering machine

62‧‧‧ Governor

7‧‧‧Parametric feedback database

71‧‧‧Parameter feedback data

72‧‧‧Parameter adjustment data

8‧‧‧ Remote Control Module

81‧‧‧Communication interface

91~96‧‧‧Steps

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a perspective view showing a conventional automatic spray system and method thereof according to the Japanese Patent Publication No. 201519954; FIG. 2 is a block diagram showing a first preferred embodiment of the intelligent scheduling system of the present invention; BRIEF DESCRIPTION OF THE DRAWINGS FIG. 4 is a block diagram showing a second preferred embodiment of the intelligent scheduling system of the present invention; and FIG. 5 is a schematic diagram of a step A flow chart illustrating a second preferred embodiment of the intelligent scheduling method of the present invention.

The detailed description of the preferred embodiments of the present invention will be apparent from the detailed description of the preferred embodiments.

Referring to FIG. 2, it is a first preferred embodiment of the intelligent scheduling system and the method thereof. The intelligent scheduling system is suitable for watering a crop 2, and includes an instant data 3, a historical database 4 An analysis module 5, an automatic watering module 6, a parameter feedback database 7, and a remote control module 8.

The real-time data 3 is a plurality of instant external environmental variable data 31 of the crop 2. In the first preferred embodiment, the crop 2 is located in a greenhouse, and the immediate external environmental variable data 31 records environmental information such as temperature, humidity, atmospheric pressure, soil, and illuminance of the environment in which the crop 2 is located. Information such as lights, fans, and air-conditioning is the control device for the greenhouse environment.

The historical database 4 stores a plurality of historical growth data 41 of the crop 2, wherein the historical growth data 41 records a pre-scheduled watering application process for the crop 2, and the crop applies information such as fertilization. In the process of watering, water washing, etc., the crop scrapping information is the scrap quantity, type, etc., the crop growth rate and shipment volume, and the historical immediate external environmental variable data 31. Through the detailed historical growth data 41, a detailed record of all the work done on the crop 2 is conducive to subsequent data analysis.

The analysis module 5 is connected to the real-time data 3 and the historical database 4, and stores a plurality of watering schedule parameter data prepared by processing the real-time external environment variable data 31 and the historical growth data 41. 51, and the watering schedule parameter data 51 records the watering time and the watering formula.

The automatic watering module 6 is connected to the analysis module 5 and includes an automatic watering machine capable of receiving the watering schedule parameter data 51. Wherein, the automatic watering machine 61 can control the controllable parameters such as the water pressure speed, the actuation speed, the flow rate and the like for the irrigation of the crop 2.

The parameter feedback database 7 is connected to the analysis module 5 and the automatic watering module 6, and stores a plurality of parameter feedback materials 71 of the watering operation process of the crop 2 by the automatic watering machine 61.

The remote control module 8 includes a communication interface 81 connected to the analysis module 5. The communication interface 81 can be a smart phone or a tablet computer, etc., except that the watering schedule can be transmitted. The parameter data 51 to the automatic watering machine 61 is further available for remote monitoring by the user. In actual implementation, the communication interface 81 can also be directly disposed on the automatic watering module 6, depending on the needs of use, and should not be limited thereto.

Referring to FIG. 3, in accordance with the intelligent scheduling system described above, the intelligent scheduling method of the present invention includes a collecting step 91, a data processing step 92, an applying step 93, a parameter feedback step 94, and an empirical learning step 95.

First, the collecting step 91 is performed to collect environmental variable data in the environment outside the crop 2, and the instant external environment variable data 31 is transmitted to the analysis module 5.

Then, the data processing step 92 is performed, and the analysis module 5 performs a data processing operation on the immediate external environment variable data 31 of the crop 2 and the historical growth data 41 to obtain the watering schedule parameter data 51 and transmit the data. To the automatic watering machine 61. Here, the watering schedule parameter data 51 records the optimal soil (media) humidity prediction, and the watering period prediction and the like.

Wherein, in the data processing step 92, the data filtering, data analysis, data normalization, data training, and data collection operations are performed on the data. In addition, the watering schedule parameter data 51 is transmitted to the automatic watering machine 61 through the communication interface 81.

It is worth mentioning that, in the collection step 91, in the process of collecting environmental variables in the environment outside the crop 2, the quality of the collected data will be affected by the on-site network environment, if the connection environment In case of poor performance, the received data will be missing. By using the data processing technology of step 92, the missing part of the data can be found and an estimated value is added, that is, the historical growth data is utilized. The data of 41 is used as a reference to find out the part of the abnormal value, and then use the data of several times before and after to fill in the estimated value.

The present invention utilizes the instant external environment variable data 31, and the historical growth data 41 to analyze the watering cycle of the greenhouse, and supplements the weathering related information to advance or delay the watering cycle. Further, the historical growth data 41 analyzes the relationship between the scrap type, the scrapped quantity, the breeding rate and the medium humidity, finds the most suitable medium humidity range of the greenhouse, and uses the prior algorithm to check the breeding rate and the highest and lowest humidity. The relationship between the two to find the high and low humidity that can produce the highest breeding rate.

Further, the watering cycle is a neural network using a nonlinear autoregressive exogenous model (NARX) architecture, and the network is trained by mini-batch and backpropagation, and the output is treated as the next time. The predicted input y _{t +1} = f ( y _t , y _{t -1} ,..., y _{t -n} , u _t ,..., u _{t -n} )+ e .

Then, the applying step 93 is performed. After the automatic watering machine 61 receives the watering schedule parameter data 51, the automatic watering machine 61 performs the watering operation on the crop 2 according to the watering schedule parameter data 51. With the automatic watering module 6, the amount of manpower used, the amount of excess water consumption can be reduced, and the problem of uneven distribution of water caused by human factors can be reduced to improve the stability of the growth of the crop 2.

Then, the parameter feedback step 94 is performed. After the automatic watering machine 61 performs the watering operation on the crop 2, the parameter feedback data 71 is generated and transmitted to the parameter feedback database 7.

Finally, in the empirical learning step 95, the parameter feedback data 71 is transmitted to the data processing operation of the data processing step 92 in the analysis module 5 as an empirical learning parameter for the next watering application.

For example, when the crop 2 is to be poured, the analysis module 5 is obtained from the historical growth data 41 of the historical database 4, and after 7 days of watering, the watering formula A is used with a water amount of 360 L, and then matched. The instant external environmental variable data 31 is obtained after the watering time is 6 days, and the watering amount is 350L. Therefore, the watering time of the watering scheduling parameter data 51 is set to 6 days, and the watering amount is 350L, and the communication interface 81 is used to The watering schedule parameter data 51 is transmitted to the automatic watering machine 61. After 6 days, the automatic watering machine 61 is automatically started to perform the watering operation on the crop 2, and the final parameter feedback data 71 is transmitted to the analysis module. The data processing operation is carried out in 5 as the empirical learning parameter for the next watering application.

The present invention uses a line-learning method to train a multi-layer perceptron to simulate a person's adjustment of watering application data, and to use a steep gradient descent to obtain an optimal watering application parameter. .

Systematize the experience of the experts, collect the real-time external environmental variable data 31, and process the historical growth data 41 to obtain the watering schedule parameter data 51, and establish a standardized implementation strategy to reduce artificial The impact of factors and the consumption of resources not only help to improve the stability of the growth of the crop 2, but also further enhance the acceptance of the industry.

The second preferred embodiment of the present invention is substantially the same as the first preferred embodiment, and the same is not described herein again. The difference is that the automatic watering module 6 further includes a controller 62 disposed on the automatic watering machine 61. The parameter feedback database 7 further stores a plurality of parameter adjustment data 72, and the intelligent scheduling method is further A manual adjustment step 96 is provided between the data processing step 92 and the applying step 93.

The controller 62 is used for a user to manually adjust the operation of the automatic watering machine 61, and the plurality of parameter adjustment data 72 is used to store the operating parameter of the controller 62. In the process, the user can refer to the The immediate external environmental variable data 31 and the parameters of the historical growth data 41 are adjusted to improve the immediacy of the watering operation.

In the manual adjustment step 96, when the automatic watering machine 61 receives the watering schedule parameter 51 and has not yet watered the crop 2, the user can manually adjust the regulator 62 to change the The automatic watering machine 61 performs the watering parameter of the crop 2, and prepares the parameter adjustment data 72, stores it in the parameter feedback database 7, and transmits the parameter adjustment data 72 to the analysis module 5. The data processing operation of the data processing step 92.

In actual implementation, after the original watering time of the watering schedule parameter 51 sent to the automatic watering machine 61 is 6 days, the water quantity is 350L, and the user can adjust the controller 62 according to his own experience to adjust the watering time. After 5 days, the water quantity is 320L, and after 5 days, the automatic watering machine 61 is automatically started to perform the watering operation on the crop 2, and the final parameter feedback data 71 is sent to the analysis module 5 for data processing. Homework, as an empirical learning parameter for the next watering application.

The parameterized feedback parameter 71 made by the instant external environment variable data 31 and the historical growth data 41, the parameter feedback data 71 made by the crop 2, and the user manually adjusting the controller 62 The parameter adjustment data 72 is made, and the analysis module 5 is fed back to the analysis module 5 for machine learning, and as a learning parameter for the next watering application, a set of standard implementation strategy is established, so that the application can be reduced without relying on the help of the expert. The amount of crops 2 produced after the completion of the crop, and at the same time increase the rate of planting.

It can be seen from the description of the above preferred embodiments that the intelligent scheduling system and the method of the present invention do have the following enhancements:

First, improve the breeding rate

With the automatic watering module 6, the amount of manpower used, the amount of excess water consumption can be reduced, and the problem of uneven distribution of water caused by human factors can be reduced to improve the stability and growth rate of the crop 2 growth.

Second, establish a standardized implementation strategy

Through the detailed historical growth data, 41 detailed records of all the work done on the crop 2 and the external environment are conducive to subsequent data analysis. Further, the expert experience is systematically collected, and the instant external environment variable data 31 is collected, and the historical growth data 41 is processed to obtain the watering schedule parameter data 51, and a standardized implementation strategy is established. To enhance the acceptance of the industry.

Third, machine learning

The instant external environment variable data 31, the historical growth data 41, the watering schedule parameter data 51, the parameter feedback data 71, and the parameter adjustment data 72 are fed back to the analysis module 5 for machine learning, and As an empirical learning parameter for the next irrigation, it can achieve automated intelligent agriculture without relying on experts.

In summary, the intelligent scheduling system and method thereof of the present invention, by the collecting step 91, the data processing step 92, the applying step 93, the parameter feedback step 94, the experience learning step 95, and the manual adjustment step 96 Between the mutual setting, the current external environment variable data 31 and the historical scheduling data 51, the watering schedule parameter data 51, the crop 2 is subjected to the parameter feedback information 71 after the watering operation, and the user manually adjusts the The parameter adjustment data 72 made by the controller 62 is fed back to the analysis module 5 for machine learning, and is used as an empirical learning parameter for the next watering application, and a set of standard implementation strategies is established, so that no need to rely on the expert's help, that is, The automatic watering can be applied to agriculture, and at the same time, the cultivation rate is increased, so that the object of the present invention can be achieved.

However, the above is only the two preferred embodiments of the present invention, and the scope of the present invention is not limited thereto, that is, the simple equivalent change of the patent application scope and the description of the invention is Modifications are still within the scope of the invention.

Claims (10)

 An intelligent scheduling system is suitable for watering a crop, comprising: an instant data, a plurality of real-time external environmental variables of the crop; a historical database storing a plurality of historical growth of the crop Data; an analysis module, connected to the real-time data and the historical database, and storing a plurality of watering schedule parameter data prepared by processing the real-time external environmental variable data and the historical growth data; An automatic watering module is connected to the analysis module, and includes an automatic watering machine capable of receiving the watering schedule parameter data; and a parameter feedback database connected to the analysis module and the automatic watering module, and storing the same There are a plurality of parameters of the automatic watering machine for the irrigation process of the crops.    The intelligent scheduling system according to claim 1, further comprising a remote control module, comprising a communication interface connected to the analysis module, configured to transmit the watering schedule parameter data to the automatic watering machine.    According to the intelligent scheduling system of claim 2, the automatic watering module further includes a controller disposed on the automatic watering machine for manually adjusting the operation of the automatic watering machine by a user. And the parameter feedback database further stores a plurality of parameter adjustment data of the controller.    According to the intelligent scheduling system described in claim 3, wherein the immediate external environmental variable data records the temperature, humidity, atmospheric pressure, soil, lighting, fan, and air-conditioning of the environment in which the crop is located. .    According to the intelligent scheduling system of claim 4, wherein the historical growth data record has a pre-scheduled watering application process for the crop, the crop application information, the crop scrap information, the crop Breeding rate, and historical data on real-time external environmental variables.    A method of the intelligent scheduling system according to any one of the preceding items 1 to 5, comprising the steps of: collecting steps, collecting environmental variable data in the environment outside the crop, and making the immediate external environment variable Data is transmitted to the analysis module; a data processing step, the analysis module performs data processing operations on the immediate external environmental variable data of the crop and the historical growth data to obtain the watering schedule parameter data, and transmits the data to the watering schedule parameter The automatic watering machine; a step of applying, after the automatic watering machine receives the watering schedule parameter data, the automatic watering machine performs watering operation on the crop according to the watering schedule parameter data; and a parameter feedback step After the automatic watering machine performs watering on the crop, the parameter feedback data is sent and transmitted to the parameter feedback database.    The intelligent scheduling method according to claim 6, wherein in the data processing step, the watering schedule parameter data is transmitted to the automatic watering machine by using the communication interface.    According to the intelligent scheduling method described in claim 7, the method further includes an empirical learning step after the parameter feedback step, in which the parameter feedback data is transmitted to the analysis module. Data processing operations for data processing steps.    The intelligent scheduling method according to claim 8 further includes a manual adjustment step between the data processing step and the applying step, in the manual adjusting step, when the automatic watering machine receives the watering row The parameter data, and before the irrigation of the crop has been applied, the user can manually adjust the regulator to change the watering parameters of the automatic irrigation machine for watering the crop, and prepare the parameter adjustment data, and then The parameter adjustment data is transmitted to the data processing operation of the data processing step in the analysis module.    According to the intelligent scheduling method described in claim 9 of the patent application, in the data processing step, the data filtering, data analysis, data normalization, data training, and data collection and the like are performed on the data.