CN113175794A - Intelligent agricultural product drying control method based on BP neural network - Google Patents
Intelligent agricultural product drying control method based on BP neural network Download PDFInfo
- Publication number
- CN113175794A CN113175794A CN202110450255.8A CN202110450255A CN113175794A CN 113175794 A CN113175794 A CN 113175794A CN 202110450255 A CN202110450255 A CN 202110450255A CN 113175794 A CN113175794 A CN 113175794A
- Authority
- CN
- China
- Prior art keywords
- drying
- neural network
- scheme
- agricultural products
- agricultural product
- 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
- 238000001035 drying Methods 0.000 title claims abstract description 231
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 80
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000004364 calculation method Methods 0.000 claims abstract description 12
- 230000000694 effects Effects 0.000 claims abstract description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 14
- 230000002452 interceptive effect Effects 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 10
- 238000010981 drying operation Methods 0.000 claims description 7
- 239000000463 material Substances 0.000 claims description 7
- 230000004913 activation Effects 0.000 claims description 5
- 239000000843 powder Substances 0.000 claims description 4
- 210000002364 input neuron Anatomy 0.000 claims description 3
- 238000005303 weighing Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 8
- 235000001674 Agaricus brunnescens Nutrition 0.000 description 7
- 238000006243 chemical reaction Methods 0.000 description 6
- 210000005069 ears Anatomy 0.000 description 5
- 102100026205 1-phosphatidylinositol 4,5-bisphosphate phosphodiesterase gamma-1 Human genes 0.000 description 2
- 206010019345 Heat stroke Diseases 0.000 description 2
- 101000691599 Homo sapiens 1-phosphatidylinositol 4,5-bisphosphate phosphodiesterase gamma-1 Proteins 0.000 description 2
- 208000007180 Sunstroke Diseases 0.000 description 2
- 235000013372 meat Nutrition 0.000 description 2
- 241000894007 species Species 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 241000233866 Fungi Species 0.000 description 1
- 241000238631 Hexapoda Species 0.000 description 1
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 241000607479 Yersinia pestis Species 0.000 description 1
- 240000008042 Zea mays Species 0.000 description 1
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 1
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 1
- 230000003187 abdominal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 235000005822 corn Nutrition 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 235000015097 nutrients Nutrition 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F26—DRYING
- F26B—DRYING SOLID MATERIALS OR OBJECTS BY REMOVING LIQUID THEREFROM
- F26B9/00—Machines or apparatus for drying solid materials or objects at rest or with only local agitation; Domestic airing cupboards
- F26B9/10—Machines or apparatus for drying solid materials or objects at rest or with only local agitation; Domestic airing cupboards in the open air; in pans or tables in rooms; Drying stacks of loose material on floors which may be covered, e.g. by a roof
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F26—DRYING
- F26B—DRYING SOLID MATERIALS OR OBJECTS BY REMOVING LIQUID THEREFROM
- F26B21/00—Arrangements or duct systems, e.g. in combination with pallet boxes, for supplying and controlling air or gases for drying solid materials or objects
- F26B21/06—Controlling, e.g. regulating, parameters of gas supply
- F26B21/08—Humidity
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F26—DRYING
- F26B—DRYING SOLID MATERIALS OR OBJECTS BY REMOVING LIQUID THEREFROM
- F26B21/00—Arrangements or duct systems, e.g. in combination with pallet boxes, for supplying and controlling air or gases for drying solid materials or objects
- F26B21/06—Controlling, e.g. regulating, parameters of gas supply
- F26B21/10—Temperature; Pressure
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F26—DRYING
- F26B—DRYING SOLID MATERIALS OR OBJECTS BY REMOVING LIQUID THEREFROM
- F26B25/00—Details of general application not covered by group F26B21/00 or F26B23/00
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F26—DRYING
- F26B—DRYING SOLID MATERIALS OR OBJECTS BY REMOVING LIQUID THEREFROM
- F26B25/00—Details of general application not covered by group F26B21/00 or F26B23/00
- F26B25/22—Controlling the drying process in dependence on liquid content of solid materials or objects
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Drying Of Solid Materials (AREA)
Abstract
The invention provides an intelligent drying control method for agricultural products based on a BP neural network, which comprises the following steps of; s1, reasoning estimated values of a plurality of characteristics of the agricultural products to obtain a characteristic combination, searching a basic drying scheme in a database according to the types of the agricultural products, and further searching a corresponding combined drying gain scheme according to the characteristic combination; s2, normalizing the characteristic combination and the expected drying degree of the agricultural products to form an input value of a BP neural network, and obtaining a BP drying gain scheme through predictive calculation; step S3, adjusting the basic drying scheme by the BP drying gain scheme and the combined drying gain scheme to form a final drying scheme; step S4, after the agricultural products are dried according to the final drying scheme, the deviation amount of the drying effect is fed back to optimize the network weight value of the agricultural products to improve the prediction accuracy of the agricultural products; the invention can get rid of excessive dependence on manual experience when drying crops through reasoning and prediction of the neural network, is convenient to operate and is easy to popularize in the basic level.
Description
Technical Field
The invention relates to the technical field of agricultural product processing, in particular to an intelligent drying control method for agricultural products based on a BP neural network.
Background
China is a big country of agricultural products, and many agricultural products such as agaric, mushroom, rice, corn and the like are often dried. The drying of agricultural products is a complex process of change, which involves many physical, chemical and biochemical changes. In the drying process, the drying quality of agricultural products is affected by different batches, different water contents of crops of different types, improper environment temperature, improper environment humidity, overlong or overlong time and other factors. For example, when the drying temperature of the mushrooms is too high, the colors of the mushrooms are blackened, nutrient substances in the mushrooms are lost, but when the temperature is too low, the mushrooms are not beneficial to being dried, the drying time of the mushrooms is prolonged, and energy waste is caused. When the wind speed is too high in the drying process, the surface of the mushroom loses water too fast and air holes are closed, so that the moisture is not discharged easily. The unreasonable drying process not only wastes energy and time, but also reduces the quality and grade of agricultural products, so that the drying quality of the agricultural products is difficult to guarantee. Seriously affecting the commodity value of agricultural products. Therefore, before drying begins, a proper drying scheme should be selected for the agricultural products.
The defects of the prior art scheme are that the temperature, the humidity and the drying time of each area are set mainly through manual experience at present, then drying is carried out according to the corresponding temperature, humidity and time setting, and drying parameters cannot be intelligently recommended according to different agricultural products.
The existing method is mainly based on manual judgment. The proper drying scheme is judged manually according to the initial state characteristics of the relevant agricultural products such as the types and appearances of the agricultural products, and even if the people have a large amount of relevant field knowledge and a large amount of relevant working experience, the judgment accuracy is difficult to ensure. Meanwhile, the artificial judgment mode which needs a great deal of knowledge and reserves experience is difficult to popularize to the basic level. It is not feasible in both accuracy and popularity.
Disclosure of Invention
The invention provides an intelligent agricultural product drying control method based on a BP neural network, which can get rid of excessive dependence on manual experience when crops are dried through reasoning and prediction of the neural network, is convenient to operate and is easy to popularize in the basic level.
The invention adopts the following technical scheme.
An intelligent agricultural product drying control method based on a BP neural network is based on a drying control system and comprises the following steps;
s1, carrying out reasoning calculation on estimated values of a plurality of characteristics of the agricultural products to be dried by using a fuzzy set formula of a fuzzy algorithm to obtain a characteristic combination which can be used for retrieval, retrieving a basic drying scheme in a database according to the types of the agricultural products, and further retrieving a corresponding combined drying gain scheme according to the characteristic combination;
step S2, normalizing the expression values of the characteristic combination and the expected drying degree of the agricultural products to form an input value of a BP neural network, and obtaining a BP drying gain scheme through the prediction calculation of the BP neural network;
step S3, adjusting a basic drying scheme by a BP drying gain scheme and a combined drying gain scheme to obtain a final drying scheme, wherein the final drying scheme comprises drying time interval division and temperature setting, humidity setting and time length setting of each drying time interval;
and step S4, after the agricultural products are dried according to the final drying scheme, feeding back a deviation value of the drying effect to the drying control system, wherein the deviation value is used as an error value of backward operation of the BP neural network, and can be used for BP neural network training to optimize a network weight value of the BP neural network to improve the prediction accuracy of the BP neural network.
In step S1, inference calculation is carried out on the estimated values of the four characteristics of the agricultural products to be dried by using a fuzzy set formula, wherein the fuzzy set is divided into three; in step S2, the BP neural network includes an input layer, a hidden layer and an output layer, and has a neural network structure of 3x4x3, including spare input neurons with a constant value of 1.
In step S2, the activation function of the BP neural network defaults to a Tanh function, which is formulated as
The four characteristics are characteristics that can be used to estimate the moisture content of the agricultural product; when the agricultural product to be dried is the agaric, the four characteristics corresponding to the agaric are the hardness, the spore powder content, the thickness of each piece and the rotten ear degree.
When the four characteristics are input in the step S1, firstly, the type of the agricultural product to be dried is input in the drying control system, and then the drying control system automatically retrieves the corresponding characteristic to be input according to the type of the agricultural product to form an interactive interface for the user to enter;
and after the basic drying scheme and the combined drying gain scheme are retrieved from the database, the interactive interface displays the retrieval result and the drying parameter setting and reasoning process corresponding to the schemes.
In the step S4, the specific operation is to feed back the drying degree of the agricultural products after the operation according to the final drying scheme in the expert system of the control system; if the drying degree of the agricultural products expected to be achieved in the step S2 is different from the fed-back drying degree, the two degrees are converted into numerical values for subtraction, and the larger the difference between the degrees is, the higher the subtraction result numerical value is and the positive and negative are; the result value is used as an error value of the BP neural network reverse operation, and after the BP neural network is trained, the related network weight value can be optimized, so that the accuracy of next prediction is improved.
An intelligent agricultural product drying control system based on a BP neural network is used for the drying control method, and the control system comprises a PLC and a drying warehouse connected with the PLC and provided with a heat pump and a fan; the PLC is connected with a remote expert system through network equipment to receive a drying scheme; the fan is connected with the moisture and hot air exhausting machine ports at the two sides and the top end of the storehouse and the fresh air port at the bottom end of the storehouse; a material tray for containing agricultural products is arranged in the middle of the storehouse; and a temperature and humidity sensor for collecting and detecting real-time temperature and humidity of each part is arranged in the storehouse.
A weighing control instrument is arranged at the position of the material tray; the temperature and humidity sensor is a 16-path temperature polling instrument and a humidity instrument which are used for acquiring temperature data and humidity data in the storehouse, and the temperature data and the humidity data acquired by the temperature and humidity sensor are uploaded to the PLC through the analog input module in the form of electric signals; the PLC controls the working conditions of the fan and the heat pump through the analog output module and the digital output module according to a final drying scheme provided by an expert system of the control system.
The control system comprises an expert system comprising a database, wherein the database of the expert system is a knowledge base comprising a crop drying process table, a crop rule matching gain table, a crop basic information table and a BP neural network parameter table;
the expert system comprises a system initial setting module, a crop species selection module, a characteristic degree selection module, an inference starting module, an inference process and result module, a generated drying scheme module, a result feedback module and a network weight module;
the expert system also comprises an interactive interface which can input the name of the agricultural product, the characteristics of the agricultural product, the expected drying degree and the drying feedback content; the interactive interface automatically retrieves the corresponding agricultural product characteristics according to the input agricultural product name, and generates a corresponding agricultural product characteristic input interface according to a retrieval result;
the crop drying process table stores existing crop basic drying process data;
the crop rule matching gain table stores drying process gain data corresponding to different characteristic combinations of different agricultural products;
the stored data of the crop basic information table comprises the names of crops and corresponding related characteristics, and a drying degree value which can be selected in the drying operation;
the BP neural network parameter table is used for storing weight values of BP neural networks used by different crops.
The expert system is used for reasoning the water content of the agricultural product according to the input agricultural product characteristics and retrieving a basic crop drying process, namely a basic drying scheme, matched with the water content from the database;
when the expert system receives the input expected drying degree of the agricultural product drying operation, the BP neural network is used for predicting the adjustable part in the drying process, the basic drying scheme is adjusted according to the prediction result, the final drying scheme is formed and output to the PLC, and the PLC controls the equipment in the drying warehouse according to the scheme setting.
The basic drying scheme records parameter values of the drying process, the BP drying gain scheme and the combined drying gain scheme record correction amounts of the parameters, and the final drying scheme is obtained by adding numerical values of the BP drying gain scheme, the combined drying gain scheme and the basic drying scheme.
The invention provides an agricultural product drying scheme expert system combining fuzzy control and a BP neural network. And the method is used for generating an optimized drying scheme of the agricultural products so as to further control the drying. The method not only provides a method for ensuring the drying quality of the agricultural products for the large country of the agricultural products in China, but also provides a means for exploring a new drying scheme from the agricultural products. The problem of drying the scheme just can not judge without relevant professional field knowledge and experience is solved, and simple and easy integrated system makes the average person also can use, very big promotion the popularization degree of system. Because the system has self-learning capability, the accuracy of next prediction can be improved by correcting errors every time. Therefore, the method has high judgment accuracy.
In the scheme of the invention, an expert system can automatically analyze the corresponding characteristics by fuzzy reasoning and match with corresponding experience rules to obtain a basic drying scheme of the corresponding agricultural products only by inputting the types of the agricultural products and selecting the corresponding characteristics by a user, and then carry out positive and negative correlation correction on the temperature, the humidity and the duration of each drying time interval in the basic drying scheme by utilizing the prediction of a BP neural network so as to obtain a final drying scheme; the excessive dependence on the experience of operators is reduced, and the method is easy to popularize in the basic level.
In the invention, because the BP neural network has self-learning capability, the weight of the BP neural network can be optimized as long as the drying result is fed back to the expert system, thereby improving the accuracy of the BP neural network in the expert system for correcting the drying scheme; thereby automatically optimizing the drying operation of the agricultural products.
The expert system is convenient to operate, and can be used only by a computer with a simple human-computer interaction function, so that the expert system is very easy to popularize at a basic level.
Drawings
The invention is described in further detail below with reference to the following figures and detailed description:
FIG. 1 is a schematic workflow diagram of the present invention;
FIG. 1a is a schematic structural diagram of the BP neural network according to the present invention;
FIG. 1b is a schematic diagram of a BP neural network production drying scheme and self-learning training according to feedback data;
FIG. 1c is a schematic diagram of a PLC and drying warehouse of the present invention;
FIG. 2 is a schematic interior view of a dry store;
FIG. 3 is a schematic diagram of the control system of the present invention;
FIG. 4 is a schematic diagram of the wiring of a temperature sensor in the temperature and humidity sensor;
FIG. 5 is a schematic diagram of the wiring of a temperature and humidity meter in the temperature and humidity sensor;
FIG. 6 is a schematic diagram of the construction of the expert system of the present invention;
in the figure: 1-PLC; 2-dry storehouse; 3-a port of a moisture and hot air discharging machine; 4-material tray; 5-fresh air port; 6-expert system; 7-network devices.
Detailed Description
As shown in the figure, the intelligent agricultural product drying control method based on the BP neural network is based on a drying control system and comprises the following steps;
s1, carrying out reasoning calculation on estimated values of a plurality of characteristics of the agricultural products to be dried by using a fuzzy set formula of a fuzzy algorithm to obtain a characteristic combination which can be used for retrieval, retrieving a basic drying scheme in a database according to the types of the agricultural products, and further retrieving a corresponding combined drying gain scheme according to the characteristic combination;
step S2, normalizing the expression values of the characteristic combination and the expected drying degree of the agricultural products to form an input value of a BP neural network, and obtaining a BP drying gain scheme through the prediction calculation of the BP neural network;
step S3, adjusting a basic drying scheme by a BP drying gain scheme and a combined drying gain scheme to obtain a final drying scheme, wherein the final drying scheme comprises drying time interval division and temperature setting, humidity setting and time length setting of each drying time interval;
and step S4, after the agricultural products are dried according to the final drying scheme, feeding back a deviation value of the drying effect to the drying control system, wherein the deviation value is used as an error value of backward operation of the BP neural network, and can be used for BP neural network training to optimize a network weight value of the BP neural network to improve the prediction accuracy of the BP neural network.
In step S1, inference calculation is carried out on the estimated values of the four characteristics of the agricultural products to be dried by using a fuzzy set formula, wherein the fuzzy set is divided into three; in step S2, the BP neural network includes an input layer, a hidden layer and an output layer, and has a neural network structure of 3x4x3, including spare input neurons with a constant value of 1.
In step S2, the activation function of the BP neural network defaults to a Tanh function, which is formulated as
The four characteristics are characteristics that can be used to estimate the moisture content of the agricultural product; when the agricultural product to be dried is the agaric, the four characteristics corresponding to the agaric are the hardness, the spore powder content, the thickness of each piece and the rotten ear degree.
When the four characteristics are input in the step S1, firstly, the type of the agricultural product to be dried is input in the drying control system, and then the drying control system automatically retrieves the corresponding characteristic to be input according to the type of the agricultural product to form an interactive interface for the user to enter;
and after the basic drying scheme and the combined drying gain scheme are retrieved from the database, the interactive interface displays the retrieval result and the drying parameter setting and reasoning process corresponding to the schemes.
In the step S4, the specific operation is to feed back the drying degree of the agricultural products after the operation according to the final drying scheme in the expert system of the control system; if the drying degree of the agricultural products expected to be achieved in the step S2 is different from the fed-back drying degree, the two degrees are converted into numerical values for subtraction, and the larger the difference between the degrees is, the higher the subtraction result numerical value is and the positive and negative are; the result value is used as an error value of the BP neural network reverse operation, and after the BP neural network is trained, the related network weight value can be optimized, so that the accuracy of next prediction is improved.
An intelligent agricultural product drying control system based on a BP neural network is used for the drying control method, and comprises a PLC1 and a drying warehouse 2 which is connected with the PLC1 and is provided with a heat pump and a fan; the PLC is connected with a remote expert system 6 through a network device 7 to receive a drying scheme; the fan is connected with the moisture and hot air exhausting machine port 3 at the two sides and the top end of the storehouse and the fresh air port 5 at the bottom end of the storehouse; a material tray 4 for containing agricultural products is arranged in the middle of the storehouse; and a temperature and humidity sensor for collecting and detecting real-time temperature and humidity of each part is arranged in the storehouse.
A weighing control instrument is arranged at the position of the material tray; the temperature and humidity sensor is a 16-path temperature polling instrument and a humidity instrument which are used for acquiring temperature data and humidity data in the storehouse, and the temperature data and the humidity data acquired by the temperature and humidity sensor are uploaded to the PLC through the analog input module in the form of electric signals; the PLC controls the working conditions of the fan and the heat pump through the analog output module and the digital output module according to a final drying scheme provided by an expert system of the control system.
The control system comprises an expert system comprising a database, wherein the database of the expert system is a knowledge base comprising a crop drying process table, a crop rule matching gain table, a crop basic information table and a BP neural network parameter table;
the expert system comprises a system initial setting module, a crop species selection module, a characteristic degree selection module, an inference starting module, an inference process and result module, a generated drying scheme module, a result feedback module and a network weight module;
the expert system also comprises an interactive interface which can input the name of the agricultural product, the characteristics of the agricultural product, the expected drying degree and the drying feedback content; the interactive interface automatically retrieves the corresponding agricultural product characteristics according to the input agricultural product name, and generates a corresponding agricultural product characteristic input interface according to a retrieval result;
the crop drying process table stores existing crop basic drying process data;
the crop rule matching gain table stores drying process gain data corresponding to different characteristic combinations of different agricultural products;
the stored data of the crop basic information table comprises the names of crops and corresponding related characteristics, and a drying degree value which can be selected in the drying operation;
the BP neural network parameter table is used for storing weight values of BP neural networks used by different crops.
The expert system is used for reasoning the water content of the agricultural product according to the input agricultural product characteristics and retrieving a basic crop drying process, namely a basic drying scheme, matched with the water content from the database;
when the expert system receives the input expected drying degree of the agricultural product drying operation, the BP neural network is used for predicting the adjustable part in the drying process, the basic drying scheme is adjusted according to the prediction result, the final drying scheme is formed and output to the PLC, and the PLC controls the equipment in the drying warehouse according to the scheme setting.
In this example, the basic drying scheme records parameter values of the drying process, the BP drying gain scheme and the combined drying gain scheme record correction amounts of the parameters, and the final drying scheme is obtained by adding numerical values of the BP drying gain scheme, the combined drying gain scheme and the basic drying scheme.
Example (b):
in the embodiment, the agricultural product drying scheme expert system is mainly divided into a database rule base part, an inference process part, an agricultural product initial parameter input part, an inference result part, a drying degree feedback part and the like.
The expert system mainly judges the corresponding drying temperature, drying humidity and drying humidity according to the characteristics of agricultural products.
The fungus is described below as an example. The moisture content of agaric can be generally estimated by appearance and the like. For example, if the ear is held by hand and the sound is crisp, the ear is pricked and elastic, the ear is not broken, the water content is relatively low, the holding is silent, the hand is pricked and the hand feeling is soft, the water content is too high.
The mature agaric which is just harvested is unfolded, soft, thick in meat quality and good in elasticity, spore powder is arranged on the abdominal surface, and the water content of the agaric is also large. The agaric harvested from Qingming to sunstroke is called spring agaric, and the spring agaric is characterized by gray black color, large and thick flowers, good quality and generally high water content. The agaric harvested from sunstroke to early autumn is called as the submerged agaric, the quality of the agaric is poor, the rotten agaric is more, the water content of the agaric is not as good as that of the agaric in spring, and mainly because of more plant diseases and insect pests and high temperature in the season. The autumn ears harvested in the beginning of autumn are called as autumn ears, the flowers are slightly small, the meat quality is medium, the quality is between that of the summer ears and that of the spring ears, and the water content is relatively higher than that of the summer ears. The temperature, humidity and drying time used in the drying process are directly influenced by the initial moisture content of the agaric. Therefore, the drying temperature, humidity and time can be predicted reasonably by using the characteristics of the agricultural products. Meanwhile, a BP neural network is introduced to correct the conclusion, so that the result is more accurate.
In use, the basic information required for reasoning is first input: the type of agricultural product, the value of the corresponding characteristic, and the desired drying degree are selected. Taking the agaric as an example, if a drying scheme is to be generated, the pull-down bar in the crop type module may be clicked first, and after the agaric is selected, the feature degree selection module may automatically read the features corresponding to the agaric, and provide corresponding feature value options in the pull-down bar. Then, the input of the initial parameters is completed by selecting the drying expectation degree in the reasoning starting module.
After the inference is started by clicking, the system substitutes numerical values of 4 features into a formula of a fuzzy set (divided into 3 equally) for calculation, and calculates a feature combination with the highest reliability after induction.
The reasoning process is displayed in the reasoning process and result module, and the generated drying scheme is displayed in the generated drying scheme module. The invoked neural network weight values will be displayed in the network weight module. Basic numerical values, combined numerical values, BP gains and total count value buttons in other operation modules of the expert system are clicked to respectively and independently check the basic drying scheme, the combined drying gain scheme, the BP drying gain scheme and the final drying scheme (the superposition result of the three schemes).
And then matching a corresponding drying combination drying gain scheme according to the characteristic combination with the highest reliability, reading a basic drying scheme in a database according to the crop type, normalizing the number of the combination and the expected drying degree, and then using the normalized number as the input of a 3x4x3BP neural network (the redundant input is constantly set to be 1 and is used as a standby), wherein the activation function of the BP neural network defaults to be a Tanh function and then calculates to obtain the BP drying gain scheme.
And obtaining a final inferred drying scheme, and drying according to the scheme. Feedback drying level may be selected in the expert system. If the expected drying degree is different from the feedback drying degree, the two degrees are converted into numerical values for subtraction, and the larger the difference between the degrees is, the higher the numerical value is, and the degree is divided into positive and negative. This value will be used as the error value for the back calculation of the BP neural network. After training, the relevant network weight values can be optimized, so that the accuracy of next prediction is improved.
During feedback, the learning rate of the neural network during reverse operation can be input in a customized mode, and the initial learning rate is 0.12 in default. The activation function defaults to a Tanh function. Clicking on the new weight values in the other operation modules can check the updated weight values in the network weight module.
In this example, the expert system uses the SQL server database, which can store the updated weights.
In the example, the PLC digital output module at the dry storehouse adopts SMART-200 EM QT16, the letters of the digital quantity represent digital variables, and the values of the digital variables are only three states: 1 represents a logic 1, i.e. high; 0 represents a logic 0, i.e., low; z represents an open circuit, i.e., a high configuration. The digital output signal is the switching value signal, 1 or 0, and the digital module has 16 digital output points. 2 blocks are selected in total and used for controlling the opening and closing of the louver fan and the fan and frequency conversion.
In this example, the SMART-200 AE04 is selected as the analog input module at the dry warehouse, 4 analog outputs are provided for a single SMART-200 AE04 module, two analog signals are selected in total, and 2 voltage or current signals are provided for the PLC, wherein the signals are generally signals transmitted by a transmitter. The method is mainly used for acquiring anemoscope, hygrometer, thermometer and wind channel opening related analog quantities in a system. The analog input module selects SMART-200 AQ04, and has 4 analog output points. The frequency conversion frequency adjustment device is mainly used for frequency conversion frequency adjustment of a fan in a control system.
In this example, SMART-200 AQ04 is selected as the analog input module at the dry warehouse, and 4 analog output points are provided. The frequency conversion frequency adjustment device is mainly used for frequency conversion frequency adjustment of a fan in a control system.
In this example, the temperature polling instrument in the dry warehouse uses XMD-1216-4-T, which is composed of 16 Pt100, when the PT100 is at 0 ℃, its resistance value is 100 ohm, and its resistance value will increase at an approximately uniform rate with the temperature rise, but the relationship between them is not a simple direct ratio, and should be close to a parabola. The temperature polling instrument indirectly measures the indoor temperature by the condition of Pt100 and by the conversion of a comparison table.
Claims (10)
1. An intelligent agricultural product drying control method based on a BP neural network is characterized in that: the drying control method is based on a drying control system and comprises the following steps;
s1, carrying out reasoning calculation on estimated values of a plurality of characteristics of the agricultural products to be dried by using a fuzzy set formula of a fuzzy algorithm to obtain a characteristic combination which can be used for retrieval, retrieving a basic drying scheme in a database according to the types of the agricultural products, and further retrieving a corresponding combined drying gain scheme according to the characteristic combination;
step S2, normalizing the expression values of the characteristic combination and the expected drying degree of the agricultural products to form an input value of a BP neural network, and obtaining a BP drying gain scheme through the prediction calculation of the BP neural network;
step S3, adjusting a basic drying scheme by a BP drying gain scheme and a combined drying gain scheme to obtain a final drying scheme, wherein the final drying scheme comprises drying time interval division and temperature setting, humidity setting and time length setting of each drying time interval;
and step S4, after the agricultural products are dried according to the final drying scheme, feeding back a deviation value of the drying effect to the drying control system, wherein the deviation value is used as an error value of backward operation of the BP neural network, and can be used for BP neural network training to optimize a network weight value of the BP neural network to improve the prediction accuracy of the BP neural network.
2. The intelligent drying control method for agricultural products based on the BP neural network as claimed in claim 1, wherein: in step S1, inference calculation is carried out on the estimated values of the four characteristics of the agricultural products to be dried by using a fuzzy set formula, wherein the fuzzy set is divided into three; in step S2, the BP neural network includes an input layer, a hidden layer and an output layer, and has a neural network structure of 3x4x3, including spare input neurons with a constant value of 1.
3. The intelligent drying control method for agricultural products based on the BP neural network as claimed in claim 2, characterized in that: in step S2, the activation function of the BP neural network defaults to a Tanh function, which is formulated as
4. The intelligent drying control method for agricultural products based on the BP neural network as claimed in claim 3, wherein: the four characteristics are characteristics that can be used to estimate the moisture content of the agricultural product; when the agricultural product to be dried is the agaric, the four characteristics corresponding to the agaric are the hardness, the spore powder content, the thickness of each piece and the rotten ear degree.
5. The intelligent drying control method for agricultural products based on the BP neural network as claimed in claim 4, wherein: when the four characteristics are input in the step S1, firstly, the type of the agricultural product to be dried is input in the drying control system, and then the drying control system automatically retrieves the corresponding characteristic to be input according to the type of the agricultural product to form an interactive interface for the user to enter;
and after the basic drying scheme and the combined drying gain scheme are retrieved from the database, the interactive interface displays the retrieval result and the drying parameter setting and reasoning process corresponding to the schemes.
6. The intelligent drying control method for agricultural products based on the BP neural network as claimed in claim 3, wherein: in the step S4, the specific operation is to feed back the drying degree of the agricultural products after the operation according to the final drying scheme in the expert system of the control system; if the drying degree of the agricultural products expected to be achieved in the step S2 is different from the fed-back drying degree, the two degrees are converted into numerical values for subtraction, and the larger the difference between the degrees is, the higher the subtraction result numerical value is and the positive and negative are; the result value is used as an error value of the BP neural network reverse operation, and after the BP neural network is trained, the related network weight value can be optimized, so that the accuracy of next prediction is improved.
7. An intelligent agricultural product drying control system based on a BP neural network, which is used in the drying control method of claim 6, and is characterized in that: the control system comprises a PLC and a drying warehouse which is connected with the PLC and is provided with a heat pump and a fan; the PLC is connected with a remote expert system through network equipment to receive a drying scheme; the fan is connected with the moisture and hot air exhausting machine ports at the two sides and the top end of the storehouse and the fresh air port at the bottom end of the storehouse; a material tray for containing agricultural products is arranged in the middle of the storehouse; and a temperature and humidity sensor for collecting and detecting real-time temperature and humidity of each part is arranged in the storehouse.
8. The intelligent agricultural product drying control system based on the BP neural network as claimed in claim 7, wherein: a weighing control instrument is arranged at the position of the material tray; the temperature and humidity sensor is a 16-path temperature polling instrument and a humidity instrument which are used for acquiring temperature data and humidity data in the storehouse, and the temperature data and the humidity data acquired by the temperature and humidity sensor are uploaded to the PLC through the analog input module in the form of electric signals; the PLC controls the working conditions of the fan and the heat pump through the analog output module and the digital output module according to a final drying scheme provided by an expert system of the control system.
9. The intelligent agricultural product drying control system based on the BP neural network as claimed in claim 8, wherein: the control system comprises an expert system comprising a database, wherein the database of the expert system is a knowledge base comprising a crop drying process table, a crop rule matching gain table, a crop basic information table and a BP neural network parameter table;
the expert system comprises a system initial setting module, a crop species selection module, a characteristic degree selection module, an inference starting module, an inference process and result module, a generated drying scheme module, a result feedback module and a network weight module;
the expert system also comprises an interactive interface which can input the name of the agricultural product, the characteristics of the agricultural product, the expected drying degree and the drying feedback content; the interactive interface automatically retrieves the corresponding agricultural product characteristics according to the input agricultural product name, and generates a corresponding agricultural product characteristic input interface according to a retrieval result;
the crop drying process table stores existing crop basic drying process data;
the crop rule matching gain table stores drying process gain data corresponding to different characteristic combinations of different agricultural products;
the stored data of the crop basic information table comprises the names of crops and corresponding related characteristics, and a drying degree value which can be selected in the drying operation;
the BP neural network parameter table is used for storing weight values of BP neural networks used by different crops.
10. The intelligent agricultural product drying control system based on the BP neural network as claimed in claim 8, wherein: the expert system is used for reasoning the water content of the agricultural product according to the input agricultural product characteristics and retrieving a basic crop drying process, namely a basic drying scheme, matched with the water content from the database;
when the expert system receives the input expected drying degree of the agricultural product drying operation, the BP neural network is used for predicting the adjustable part in the drying process, the basic drying scheme is adjusted according to the prediction result, the final drying scheme is formed and output to the PLC, and the PLC controls the equipment in the drying warehouse according to the scheme setting.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110450255.8A CN113175794A (en) | 2021-04-25 | 2021-04-25 | Intelligent agricultural product drying control method based on BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110450255.8A CN113175794A (en) | 2021-04-25 | 2021-04-25 | Intelligent agricultural product drying control method based on BP neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113175794A true CN113175794A (en) | 2021-07-27 |
Family
ID=76925952
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110450255.8A Pending CN113175794A (en) | 2021-04-25 | 2021-04-25 | Intelligent agricultural product drying control method based on BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113175794A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117064031A (en) * | 2023-08-30 | 2023-11-17 | 嘉兴美旺机械制造有限公司 | Food processing baking method and system based on flowing layer |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EA200101137A1 (en) * | 1999-04-28 | 2002-04-25 | Бюлер Аг | METHOD AND DEVICE FOR OPTIMIZATION OF PROCESS MANAGEMENT, AND ALSO CONTROL OVER THE PROCESS IN INSTALLATION FOR MANUFACTURING PASTA |
| CN106227038A (en) * | 2016-07-29 | 2016-12-14 | 中国人民解放军信息工程大学 | Grain drying tower intelligent control method based on neutral net and fuzzy control |
| CN106352680A (en) * | 2016-10-28 | 2017-01-25 | 佳木斯大学 | Agaric double-section drying device and drying control method thereof |
| CN106444379A (en) * | 2016-10-10 | 2017-02-22 | 重庆科技学院 | Intelligent drying remote control method and system based on internet of things recommendation |
| CN109708459A (en) * | 2019-01-30 | 2019-05-03 | 深圳市森控科技有限公司 | A kind of intelligence drying control method, system and device |
| CN109708460A (en) * | 2019-01-30 | 2019-05-03 | 深圳市森控科技有限公司 | A kind of drying system and its control method and control device |
| CN109916149A (en) * | 2019-04-11 | 2019-06-21 | 湖北裕山菌业有限公司 | Dryer |
-
2021
- 2021-04-25 CN CN202110450255.8A patent/CN113175794A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EA200101137A1 (en) * | 1999-04-28 | 2002-04-25 | Бюлер Аг | METHOD AND DEVICE FOR OPTIMIZATION OF PROCESS MANAGEMENT, AND ALSO CONTROL OVER THE PROCESS IN INSTALLATION FOR MANUFACTURING PASTA |
| CN106227038A (en) * | 2016-07-29 | 2016-12-14 | 中国人民解放军信息工程大学 | Grain drying tower intelligent control method based on neutral net and fuzzy control |
| CN106444379A (en) * | 2016-10-10 | 2017-02-22 | 重庆科技学院 | Intelligent drying remote control method and system based on internet of things recommendation |
| CN106352680A (en) * | 2016-10-28 | 2017-01-25 | 佳木斯大学 | Agaric double-section drying device and drying control method thereof |
| CN109708459A (en) * | 2019-01-30 | 2019-05-03 | 深圳市森控科技有限公司 | A kind of intelligence drying control method, system and device |
| CN109708460A (en) * | 2019-01-30 | 2019-05-03 | 深圳市森控科技有限公司 | A kind of drying system and its control method and control device |
| CN109916149A (en) * | 2019-04-11 | 2019-06-21 | 湖北裕山菌业有限公司 | Dryer |
Non-Patent Citations (2)
| Title |
|---|
| 吕伟: "农副产品智能化干燥系统的分析与设计", 《中国优秀博硕士学位论文全文数据库(硕士) 工程科技Ⅱ辑》 * |
| 吕伟等: "BP神经网络PID控制算法在农作物干燥控制系统中的应用研究与设计", 《计算机测量与控制》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117064031A (en) * | 2023-08-30 | 2023-11-17 | 嘉兴美旺机械制造有限公司 | Food processing baking method and system based on flowing layer |
| CN117064031B (en) * | 2023-08-30 | 2024-04-19 | 嘉兴美旺机械制造有限公司 | Food processing baking method and system based on flowing layer |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ruan et al. | Prediction of dough rheological properties using neural networks | |
| CN106472412B (en) | pet feeding method and system based on internet of things | |
| CN110119766B (en) | Green pepper greenhouse temperature intelligent early warning device of multi-combination intelligent model | |
| CN110119169B (en) | Tomato greenhouse temperature intelligent early warning system based on minimum vector machine | |
| CN109634098A (en) | A kind of fattening house environment conditioning system and method | |
| CN106614273B (en) | Pet feeding method and system based on Internet of Things big data analysis | |
| CN106472332A (en) | Pet feeding method and system based on dynamic intelligent algorithm | |
| CN110069032B (en) | Eggplant greenhouse environment intelligent detection system based on wavelet neural network | |
| CN108739052A (en) | A kind of system and method for edible fungi growth parameter optimization | |
| CN110109193A (en) | A kind of eggplant greenhouse temperature intellectualized detection device based on DRNN neural network | |
| CN110119767B (en) | Intelligent cucumber greenhouse temperature detection device based on LVQ neural network | |
| CN113175794A (en) | Intelligent agricultural product drying control method based on BP neural network | |
| CN113126490A (en) | Intelligent frequency conversion oxygenation control method and device | |
| Irawan et al. | Intelligent Irrigation Water Requirement System Based on Artificial Neural Networks and Profit Optimization for Planning Time Decision Making of Crops in Lombok Island | |
| CN111553806B (en) | Self-adaptive crop management system and method based on low-power-consumption sensor and Boost model | |
| CN215063282U (en) | Intelligent agricultural product drying control system based on BP neural network | |
| CN117354988A (en) | Greenhouse LED illumination closed-loop control method and system based on ambient light sensing | |
| KR102387765B1 (en) | Method and apparatus for estimating crop growth quantity | |
| CN115959933B (en) | Intelligent control method and system for aerobic composting | |
| Chen et al. | A water-saving irrigation decision-making model for greenhouse tomatoes based on genetic optimization TS fuzzy neural network | |
| Mimboro et al. | Weather monitoring system AIoT based for oil palm plantation using recurrent neural network algorithm | |
| Tripathy et al. | Smart Farming based on Deep Learning Approaches | |
| TW202018650A (en) | Intelligent crop growth optimization system automatically generating and building an optimized sequence command that incorporates a dynamic change technology and a dynamic replication technology | |
| Hadi et al. | Mamdani fuzzy logic-based smart measuring device as quality determination for grain post-harvest technology | |
| TW202025063A (en) | A method for calculating a growth stage of a crop and computer program product |
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: 20210727 |