Method and system for regulating the internal environment of a greenhouse for crop cultivation

Abstract

The present application relates to the technical field of greenhouse internal environment regulation, solves the technical problem that the internal environment of the greenhouse cannot be accurately regulated, resulting in high management cost, and particularly relates to a method for regulating the internal environment of a greenhouse for crop planting, comprising: obtaining current environment data of an agricultural greenhouse at a current time collected by a data collection device; inputting the current environment data into an environment prediction model and obtaining the evolution rule of the internal environment data of the agricultural greenhouse based on time according to the output of the environment prediction model; constructing a fitness function of a genetic algorithm according to the evolution rule of the internal environment data of the agricultural greenhouse based on time; and solving the optimal regulation strategy for the internal environment of the agricultural greenhouse according to the current environment data and the fitness function. The present application uses a compensation fuzzy neural network optimized based on a genetic algorithm, comprehensively considers the constraint relationship between various environment variables in the greenhouse, obtains an optimal regulation strategy, accurately regulates the environment parameters, and reduces the management cost.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of greenhouse internal environment regulation, and particularly relates to a method and system for regulating the internal environment of a greenhouse for crop planting. BACKGROUND

[0002] In recent years, with the continuous development of agricultural technology, agricultural greenhouse applications are becoming more and more widespread. In order to avoid the problems of low crop yield and high management cost caused by human error, intelligent regulation methods for the internal environment of agricultural greenhouses based on neural networks have appeared on the market.

[0003] For example, a Chinese patent application No. 201711423181.9 discloses a greenhouse intelligent management method and system. The method constructs an environment model based on a neural network according to the environmental information in the main greenhouse, processes and analyzes the environmental information input into the environment model to obtain a control instruction for adjusting the environmental parameters in the main greenhouse, and sends the control instruction to the control device of the slave greenhouse to meet the individual environmental needs of different crops. However, in the above technical solution, since the environmental parameters of the main greenhouse and the slave greenhouse are not the same, sending the control instruction obtained from the environmental information of the main greenhouse to the slave greenhouse cannot achieve accurate regulation of the environmental parameters of the slave greenhouse.

[0004] Moreover, most of the current regulation methods ignore the fact that various environmental variables inside the agricultural greenhouse change over time and affect each other. Therefore, only processing and analyzing the collected environmental information based on the environment model constructed by the neural network cannot obtain the optimal regulation strategy, which easily causes energy waste and is not conducive to the growth of crops in the greenhouse, thereby increasing the management cost of the greenhouse. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a method and system for regulating the internal environment of a greenhouse for crop planting, which solves the technical problem of being unable to accurately regulate the internal environment of the greenhouse, resulting in energy waste and increasing the management cost of the greenhouse, and achieves the purposes of accurately regulating the internal environment of the agricultural greenhouse, saving energy and reducing management cost.

[0006] To solve the above technical problems, the present application provides the following technical solution: a method for regulating the internal environment of a greenhouse for crop planting, the regulation method comprising the following steps:

[0007] S1, obtaining current environmental data of the agricultural greenhouse at the current time collected by a data collection device;

[0008] S2, inputting the current environmental data into an environment prediction model to obtain the time-based evolution rule of the internal environmental data of the agricultural greenhouse;

[0009] S3, constructing a fitness function of the genetic algorithm based on the time-based evolution rule of the agricultural greenhouse internal environment data;

[0010] S4, initializing the initial population number of the genetic algorithm according to the current environment data, and encoding the initial population;

[0011] S5, repeatedly selecting, crossing and mutating the initial population in sequence according to the fitness function to solve the optimal control strategy of the agricultural greenhouse internal environment;

[0012] S6, obtaining a control instruction applied to the internal environment parameters of the greenhouse according to the optimal control strategy;

[0013] S7, sending the control instruction to the corresponding execution device to control the internal environment of the greenhouse through the execution device.

[0014] Further, in step S2, the current environment data is input into the environment prediction model to obtain the time-based evolution rule of the agricultural greenhouse internal environment data, and the specific process includes the following steps:

[0015] S201, inputting the current environment data into the environment prediction model to initialize the network parameters;

[0016] S202, presetting the fuzzy partition number of the environment prediction model as 2, the learning rate as 1.0, the compensation degree as 0.8, and the expected error as 0.001, and taking the initialized parameters as the starting point to perform the compensation fuzzy reasoning and error back propagation process, and iterating until the predetermined expected error is reached;

[0017] S203, the environment prediction model after the iterative process determines the environment data y of the agricultural greenhouse internal environment to the next monitoring cycle according to the current environment data, and the environment data y is the time-based evolution rule of the agricultural greenhouse internal environment data, and the expression of the environment data y is:

[0018]

[0019] In the above formula, y(x i ) represents the output variable of the environment prediction model, ω i represents the connection weight between the i-th input layer and the output layer, x i represents the input variable of the environment prediction model, f(x i ) represents the membership function of each input variable, and N1 represents the number of nodes of the input layer.

[0020] Further, in step S3, x(T) represents the current time greenhouse internal environment temperature, y(T) represents the predicted greenhouse internal environment temperature of the environment prediction model to the next monitoring cycle, and the expression of the fitness function is:

[0021] If y(T) > x(T), it indicates that the air temperature of the greenhouse interior environment gradually increases from the current time to the next monitoring period, and the expression of the fitness function for the management of the greenhouse interior environment in this process is:

[0022]

[0023] If y(T) ≤ x(T), it indicates that the air temperature of the greenhouse interior environment gradually decreases from the current time to the next monitoring period, and the expression of the fitness function for the management of the greenhouse interior environment in this process is:

[0024]

[0025] In the above formula, represents the average value of the variable x i , p represents the number of output variables y(x i ) of the environment prediction model, and y(x i ) max represents the maximum value of the variable x i , and α takes a value in the range of [0, 1], and takes a value of 0.5.

[0026] Further, in step S4, the expression for encoding the initial population is:

[0027] Z = [(α1, β1), (α2, β2)…(α i , β i )…(α M , β M )]

[0028] In the above formula, Z represents a control strategy of the greenhouse environment, α i , β i respectively represent the control amount and the control duration of the i-th environmental parameter, i = 1, 2, …, P, and P represents the type of environmental parameters contained in the current environmental data.

[0029] Further, in step S5, the method for solving the optimal control strategy of the interior environment of the agricultural greenhouse is as follows:

[0030] S51, solve the objective function of the control strategy corresponding to each individual in the population;

[0031] The expression of the objective function is:

[0032]

[0033] In the above formula, S represents the total amount of energy consumed by the control of the greenhouse interior environment, P represents the type of environmental parameters contained in the current environmental data, and Q i,lrepresents the average amount of energy consumption of the adjusting device for adjusting the i-th environmental parameter in the l-th time period, T represents the length of the greenhouse environment adjusting time period;

[0034] S52, fast non-dominated sorting and crowdedness calculation are performed on the initial population according to the target function to determine fitness values;

[0035] S53, selection operation, crossover operation and mutation operation are performed on the initial population according to the repeated sorting of the fitness function, global evaluation decision is made, and the optimal greenhouse environment regulation strategy is output.

[0036] Further, in step S7, the regulation instruction is sent to the corresponding execution device, and the greenhouse internal environment is regulated by the execution device, which can also be set as:

[0037] Taking the time of sending the regulation instruction for the first time as the starting point, the internal environment data of the greenhouse is collected by the data collection device every interval preset time period, the post-environment data is obtained, and it is judged whether the post-environment data is located in the normal range or not;

[0038] If yes, the post-environment data is in the normal range, and the present regulation is ended;

[0039] If no, return to step S4 for circulation.

[0040] The application also provides a technical scheme: a system for realizing the regulation method of the internal environment of the greenhouse for planting crops, comprising: a control module, a data collection device, a cloud storage platform and an execution device;

[0041] The control module comprises a main controller and a plurality of field controllers, the field controller is used for processing the current environment data collected by the data collection device in the greenhouse to obtain an optimal regulation strategy, and the current environment data and the optimal regulation strategy are sent to the main controller through a network, the main controller is used for receiving the current environment data and the optimal regulation strategy sent by each field controller, and sending a regulation instruction to the corresponding field controller according to the optimal regulation strategy;

[0042] The data collection device establishes data communication with the field controller through the Internet of Things, and the data collection device is used for collecting the current environment data in the greenhouse;

[0043] The cloud storage platform establishes data communication with the field controller through a wireless network, and the cloud storage platform is used for storing the greenhouse environment data of various greenhouse crops at various stages in the whole life cycle;

[0044] The execution device establishes data communication with the field controller through the Internet of Things, and the execution device is used for adjusting the environment in the greenhouse

[0045] Further, the control module further comprises a human-computer interaction device, the human-computer interaction device is in data communication with the main controller, and the human-computer interaction device is used for user to view the greenhouse environment and perform control operation.

[0046] Further, the execution device comprises a ventilation device, a roller blind device, a water pump, a carbon dioxide generator and a light supplement lamp.

[0047] The ventilation device and the roller blind device are used for performing corresponding actions according to the regulation instruction to adjust the temperature inside the greenhouse.

[0048] The water pump is used for performing corresponding actions according to the regulation instruction to adjust the humidity inside the greenhouse.

[0049] The carbon dioxide generator is used for performing corresponding actions according to the regulation instruction to adjust the carbon dioxide concentration inside the greenhouse.

[0050] The light supplement lamp is used for performing corresponding actions according to the regulation instruction to adjust the light intensity inside the greenhouse.

[0051] By the above technical scheme, the present application provides a regulation method and system for the internal environment of a greenhouse for planting crops, which has at least the following beneficial effects:

[0052] 1. The present application considers the constraint relationship between various environmental variables inside the greenhouse from the overall aspect to obtain an optimal regulation strategy for the internal environment of the greenhouse, realizes accurate regulation of the internal environmental parameters of the greenhouse, and thus reduces the management cost of the greenhouse, thereby having high social value and application prospect.

[0053] 2. The present application collects the subsequent environmental data inside the greenhouse every interval of a preset time period starting from the time of sending the regulation instruction for the first time, and judges whether the subsequent environmental data is within the normal range. If the subsequent environmental data is not within the normal range, the regulation instruction is reacquired until the internal environmental data of the greenhouse is adjusted to be within the normal range, otherwise the present regulation is ended. The regulation strategy based on time change is generated according to the latest environmental parameters inside the greenhouse, which can effectively avoid energy waste and further reduce the management cost of the agricultural greenhouse.

[0054] 3. The present application generates corresponding regulation instructions according to the optimal regulation strategy by the main controller, sends the regulation instructions to corresponding execution devices through the field controller, and displays on the human-computer interaction device. It not only facilitates timely adjustment of the internal environment of the greenhouse, but also can warn the staff, ensures the crops in the greenhouse from multiple aspects to have a favorable growing environment, and has the advantages of stable operation, rapidness and accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0055] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

[0056] Figure 1 Flow chart of the regulation method in the embodiment one of the application;

[0057] Figure 2 Principle diagram of the compensation fuzzy neural network in the embodiment one of the application;

[0058] Figure 3 Flow chart of the regulation method in the embodiment two of the application;

[0059] Figure 4 Principle diagram of the regulation system in the embodiment three of the application.

[0060] In the figure: 100, control module; 200, data acquisition device; 300, cloud storage platform; 400, execution device. DETAILED DESCRIPTION

[0061] In order to make the above objectives, features and advantages of the application more apparent, more comprehensible and more understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. The realization process of how the application applies technical means to solve technical problems and achieve technical effects can be fully understood and implemented.

[0062] In the embodiment, the internal environmental index factors of the agricultural greenhouse mainly include temperature, humidity, light intensity and carbon dioxide concentration, wherein the temperature and humidity are the key factors affecting the growth of crops in the greenhouse. However, the temperature and humidity are two contradictory bodies, that is, when one of them is adjusted, the other one will be affected. In addition, with the change of light intensity, the temperature and humidity inside the greenhouse also change, therefore, the application proposes to use the compensation fuzzy neural network based on genetic algorithm optimization to comprehensively consider the constraint relationship between various environmental variables inside the greenhouse from the global, obtain the optimal regulation strategy for the greenhouse internal environment based on time change, realize the accurate regulation of the greenhouse internal environment parameters, and achieve the purpose of reducing the greenhouse internal environment regulation cost.

[0063] Embodiment one

[0064] Please refer to Figure 1 and Figure 2, the embodiment shows a specific embodiment, the embodiment adopts the compensation fuzzy neural network based on genetic algorithm optimization, inputs the current environment data of the agricultural greenhouse at the current time collected into the environment prediction model, obtains the time-based evolution rule of the agricultural greenhouse internal environment data, and constructs the fitness function of the genetic algorithm according to the time-based evolution rule of the agricultural greenhouse internal environment data, then initializes the initial population of the genetic algorithm according to the current environment data, and repeatedly selects, crosses and mutates the initial population according to the fitness function, considers the constraint relationship between the multiple environment variables in the greenhouse from the global, to obtain the optimal control strategy of the agricultural greenhouse internal environment, realizes the accurate control of the greenhouse internal environment parameters, thereby reduces the management cost of the greenhouse, and has high social value and application prospect.

[0065] Please refer to Figure 1 , the embodiment proposes a control method for the internal environment of a greenhouse for crop planting, as shown in Figure 1 , comprising the following steps:

[0066] S1, obtaining the current environment data of the agricultural greenhouse at the current time collected by the data acquisition device, the change of the internal environment of the agricultural greenhouse is usually affected by outdoor air, crop growth, greenhouse structure and material, greenhouse cover and various greenhouse environment control devices. Therefore, the current environment data of the agricultural greenhouse at the current time collected by the data acquisition device includes temperature value, humidity value, light intensity value and carbon dioxide concentration value.

[0067] S2, input the current environment data into the environment prediction model to obtain the time-based evolution rule of the agricultural greenhouse internal environment data. The current environment data is input into the environment prediction model constructed based on the compensation fuzzy neural network to initialize the network parameters, and the statistical analysis method is used to preset the fuzzy segmentation number as 2, the learning rate as 1.0, the compensation degree as 0.8 and the expected error as 0.001, then the compensation fuzzy reasoning and error back propagation process are carried out with the initialized parameters as the starting point, which can be iterated until the predetermined expected error is reached. After a series of dynamic global optimization operations, the environment data y of the agricultural greenhouse internal environment at the next monitoring period is output, and the specific expression is as follows:

[0068]

[0069] In the above formula, y(x i ) represents the output variable of the environment prediction model, ω i represents the connection weight between the i-th input layer and the output layer, x i represents the input variable of the environment prediction model, f(x i ) represents the membership function of each input variable, and N1 represents the number of nodes of the input layer.

[0070] It should be noted that the greenhouse internal environment data not only changes with time, but also is affected by the type and growth period of crops in the greenhouse. Therefore, when regulating the internal environment of the greenhouse, the influence of the crops on the internal environment data of the greenhouse at the current growth period with the passage of time needs to be calculated based on the input and output of the environment prediction model, that is, the time-based evolution rule of the internal environment data of the agricultural greenhouse refers to the change rule of the internal environment of the agricultural greenhouse from the current time to the next monitoring period.

[0071] Therefore, in the embodiment, the optimized environment prediction model is used to determine the environment data of the internal environment of the agricultural greenhouse to the next monitoring period. This optimization not only optimizes the compensation fuzzy operation, but also makes the optimal model obtained by training more stable, and greatly improves the fault tolerance performance of the environment prediction model, so as to more accurately simulate the dynamic change of the real environment in the agricultural greenhouse.

[0072] The dynamic evolution rule can more accurately determine the dynamic change of the internal environment parameters of the agricultural greenhouse, and obtain the change rule of the internal environment of the agricultural greenhouse from the current time to the next monitoring period. Therefore, under the premise of establishing the evolution rule, the internal environment of the greenhouse can be better regulated, thereby providing data basis for the reliability and accuracy of regulation.

[0073] Therefore, in the embodiment, the specific construction steps of the environment prediction model described in step S2 are as follows:

[0074] S21, obtain the historical environment data stored in the cloud storage platform, the historical environment data including a plurality of environment data at the current time and the collection period before and after the current time in the historical date. The data stored in the cloud storage platform includes the reference range of various variables established at each stage of the full life cycle of different varieties of crops in the agricultural greenhouse, and the values of various variables of the internal environment of the greenhouse at different times collected by the data collection device.

[0075] In the embodiment, the obtained historical data is the same as the same variety of greenhouse crops in the same growth stage and at the current time and the collection period before and after the current time, and each historical environment data includes four greenhouse environment parameters, i.e. temperature value, humidity value, carbon dioxide concentration value and light intensity value.

[0076] S22, the historical environment data is preprocessed by wavelet denoising method to obtain the preprocessed data sample set. Specifically, the data features of the noisy historical environment data are extracted, then filtered by low-pass filtering, and then reconstructed to obtain the preprocessed data sample set. In the embodiment, the wavelet denoising adopts a global unified threshold method, which is common knowledge for those skilled in the art and will not be described in detail here.

[0077] S23, training the compensation fuzzy neural network according to the data sample set to obtain a trained environment prediction model. Specifically, the 100 historical environment data with regulation labels obtained are randomly divided into a training set and a test set according to a preset ratio, wherein the training set is used to train network parameters of the compensation fuzzy neural network, and there are 80 samples in total; the test set is used to test the performance of the environment prediction model based on the compensation fuzzy neural network, and there are 20 samples in total.

[0078] Then, the 80 samples of the training set are input into the compensation fuzzy neural network to initialize the network and preset the initial value of the network parameter, and the compensation fuzzy reasoning and error back propagation process are performed from this starting point, and the loop is iterated until the predetermined expected error is reached, to obtain the trained environment prediction model. The expected error involved here is usually the absolute value of the difference between the expected output and the actual output. The compensation neural network is a hybrid system combining compensation fuzzy logic and neural network, which is composed of fuzzy neurons for control and decision-making. These fuzzy neurons are defined to perform fuzzification operation, fuzzy reasoning, compensation fuzzy operation and defuzzification operation. As shown in the figure, a compensation fuzzy neural network has a six-layer structure: input layer, fuzzification layer, fuzzy reasoning layer, compensation operation layer, defuzzification layer and output layer. The specific data processing process is as follows: Figure 2

[0079] The first layer is the input layer, which only introduces the obtained greenhouse environment data into the network without performing any information processing. Assuming that the number of nodes N1 = n of this layer, the input vector is x = [x1, x2, …, xn], each component x n ] T is a fuzzy language variable; i

[0080] The second layer is the fuzzification layer, which performs fuzzy space segmentation on each component. Each node represents a language variable value. The fuzzy set of each language variable value after the fuzzification processing of the input vector is denoted as wherein is the jth language variable value of x i , k i is the fuzzy segmentation number of x i . The role of this layer is to calculate the membership function of each input component belonging to each language variable value fuzzy set. The expression of the membership function is: c ij represents the center of the membership function, and σ ij represents the width of the membership function. The number of nodes of this layer is

[0081] ​​The third layer is a fuzzy inference layer, each node represents a fuzzy inference rule, the role of the layer is to match the fuzzy rules, and to calculate the fitness of each rule, that is: Wherein, i1=1, 2, …, k1; i2=1, 2, …, k2; …; i n =1, 2, …, k n , j=1, 2, …, N2, the number of nodes of the layer is the same as the number of nodes of the second layer;

[0082] The fourth layer is a compensation operation layer, the number of compensation fuzzy neurons is the same as the number of nodes of the fuzzy rule, the compensation fuzzy neuron can compensate a relatively compromised output between the best input x1 and the worst input x2, that is Wherein, γ∈[0, 1] is a compensation degree, the layer adopts a compensation operation based on positive operation and negative operation;

[0083] The fifth layer is a defuzzification layer, which adopts a defuzzification function to perform a defuzzification operation, each node of the layer represents an output, that is, the number of nodes is equal to the number of outputs;

[0084] The sixth layer is an output layer, which represents the fuzzy inference output by a specific method, assuming that the layer has m nodes, the output vector is y=[y1, y2, …, y m ] T .

[0085] Finally, 20 samples of the test set are input into the trained environment prediction model to initialize the model parameters, and statistical analysis method is used to preset the fuzzy segmentation number as 2, the learning rate as 1.0, the compensation degree as 0.8, and the expected error as 0.001, then, the compensation fuzzy inference and error back propagation process are performed with the initialized parameters as the starting point, which can be iterated until the predetermined expected error is reached, and the optimal environment prediction model based on the compensation fuzzy neural network is obtained.

[0086] It should be noted that in the process of training the compensation fuzzy neural network, the fuzzy operation adopts a dynamic and global optimization operation, and the compensation fuzzy operation is dynamically optimized in the neural network learning algorithm, which not only makes the optimal model obtained by training more stable, but also greatly improves the network fault tolerance performance, so as to more accurately simulate the dynamic changes of the real environment of the greenhouse.

[0087] Through the embodiment, firstly, an environment prediction model based on a compensation fuzzy neural network is constructed, a fitness function is constructed according to input and output of the environment prediction model, an initial population of a genetic algorithm is initialized according to current environment data, and selection, crossover and mutation operations are repeatedly performed on the initial population in sequence according to the fitness function, so that the optimal control strategy of the environment in the agricultural greenhouse is obtained by comprehensively considering the constraint relationship between various environment variables in the greenhouse from a global perspective, the precise control of the environment parameters in the greenhouse is realized, the management cost of the greenhouse is reduced, and the greenhouse has high social value and application prospect.

[0088] S3, constructing a fitness function of the genetic algorithm based on time evolution rules of the environment data in the agricultural greenhouse. In the embodiment, x(T) represents the temperature of the environment in the greenhouse at the current time, y(T) represents the temperature of the environment in the greenhouse predicted by the environment prediction model until the next monitoring period, and if y(T)>x(T), it indicates that the air temperature of the environment in the greenhouse gradually increases from the current time to the next monitoring period, and the expression of the fitness function for the control of the environment in the greenhouse in this process is:

[0089]

[0090] In the above formula, represents the average value of the variable x i , and a takes a value in the range of [0, 1], and in the embodiment, a takes a value of 0.5.

[0091] If y(T)≤x(T), it indicates that the air temperature of the environment in the greenhouse gradually decreases from the current time to the next monitoring period, and the expression of the fitness function for the control of the environment in the greenhouse in this process is:

[0092]

[0093] In the above formula, y(x i ) max represents the maximum value of the variable x i , p represents the number of output variables y(x i ) of the environment prediction model, a takes a value in the range of [0, 1], and in the embodiment, a takes a value of 0.5.

[0094] S4, initializing the number of initial populations of the genetic algorithm according to the current environment data, and encoding the initial population. Firstly, a population with P individuals is randomly generated according to the types P of environment parameters contained in the current environment data, and then the initial population is encoded by using a real number coding scheme, and the specific coding rule expression is:

[0095] Z = [(a1, b1), (a2, b2), …, (a M , bM ]

[0096] In the above formula, Z represents a control strategy of the greenhouse environment, a i ,β i respectively represent the control amount and the control duration of the i-th environment parameter, i = 1, 2, …, P, and P represents the type of the environment parameter contained in the current environment data.

[0097] By using the above encoding rule to express the control amount and the control duration of each environment parameter with a set of real numbers, the control strategy is very intuitive, and it is easy to accurately control the control duration of each variable, thereby improving the management accuracy of the greenhouse internal environment.

[0098] S5, repeatedly selecting, crossing and mutating the initial population according to the fitness function to solve the optimal control strategy of the agricultural greenhouse internal environment. The specific method for realizing this step includes:

[0099] S51, solving the objective function of the control strategy corresponding to each individual in the population. When adjusting the temperature value and the humidity value as the main target, and adjusting the light intensity value and the carbon dioxide concentration value as the sub-target, the objective function expression of the corresponding control strategy is as follows:

[0100]

[0101] In the above formula, S represents the total amount of energy consumed by the greenhouse internal environment control, P represents the type of the environment parameter contained in the current environment data, Q i,l represents the average amount of energy consumption of the adjusting device for adjusting the i-th environment parameter in the l-th time period, and T represents the length of the greenhouse environment adjustment time period.

[0102] S52, determining the fitness value according to the fast non-dominated sorting and the crowdedness calculation of the initial population. Specifically, the current environment data is input into the greenhouse environment model, the greenhouse environment model is learned to have the mapping rule of the output, and the output of the greenhouse environment model is taken as the fitness value.

[0103] S53, repeatedly selecting, crossing and mutating the initial population according to the fitness function to evaluate and decide from the global perspective, and output the optimal control strategy of the greenhouse environment. Specifically, the initial population is repeatedly selected, crossed and mutated, and evaluated and decided from the global perspective of the greenhouse environment adjustment, until the preset iteration number is reached, and the optimal control strategy of the greenhouse internal environment control is output.

[0104] In this embodiment, the current environment data is processed by the environment prediction model, and it is necessary to adjust the greenhouse environment. The iteration number is set to 100, and after 100 iterations, the internal environment of the agricultural greenhouse is comprehensively evaluated and decided from a global perspective. The final result is usually a group of results after mutual trade-off between multiple contradictory targets, that is, the optimal control strategy of the internal environment management of the greenhouse.

[0105] S6, obtaining a control instruction for the internal environment parameter of the greenhouse according to the optimal control strategy. The control instruction for controlling the running state of the execution device is obtained according to the optimal control strategy of the internal environment management of the greenhouse, wherein the execution device includes a ventilation device, a roller blind device, a water pump, a carbon dioxide generator and a light supplement lamp.

[0106] It should be noted that the control of the temperature of the agricultural greenhouse includes three aspects of temperature maintenance, temperature rise and temperature reduction, wherein the cooling methods include shading, water flow and forced ventilation; the control of humidity includes dehumidification and humidification, and reasonable control and adjustment of humidity can effectively prevent the crops in the greenhouse from causing diseases and pests due to excessive humidity concentration. The dehumidification technology is natural ventilation dehumidification or ventilation device dehumidification, and the humidification method is usually spray humidification and wet curtain humidification.

[0107] S7, sending the control instruction to the corresponding execution device to control the internal environment of the greenhouse. The control instruction obtained according to the optimal control strategy of the internal environment management of the greenhouse is sent to the corresponding execution device in the agricultural greenhouse, and the internal environment of the agricultural greenhouse is adjusted by the running of the execution device.

[0108] It should be noted that when the temperature inside the agricultural greenhouse is high and the light intensity is low, the optimal control strategy is to ventilate and roll up the cover curtain, and the generated control instruction is to start the ventilation device and the roller blind device and the running time length. The instructions are sent to the ventilation device and the roller blind device to start, and stop running when the preset running time length of the control instruction is reached.

[0109] Embodiment two

[0110] Please refer to Figure 3 , another specific embodiment of the present application is shown, which is realized on the basis of the control method of the internal environment of the greenhouse proposed in embodiment one. The same or similar parts are referred to each other, and have the beneficial effects of the corresponding method embodiments. The same parts will not be described here.

[0111] As Figure 3 shown, a control method of the internal environment of a greenhouse for crop planting includes the following steps:

[0112] S1, obtaining current environment data of the agricultural greenhouse at the current time collected by a data acquisition device;

[0113] S2, inputting the current environment data into an environment prediction model to obtain an environment data evolution rule of the agricultural greenhouse based on time;

[0114] S3, constructing a fitness function of a genetic algorithm according to the environment data evolution rule of the agricultural greenhouse based on time;

[0115] S4, initializing the initial population number of the genetic algorithm according to the current environment data, and encoding the initial population;

[0116] S5, repeatedly selecting, crossing and mutating the initial population according to the fitness function to solve the optimal control strategy of the internal environment of the greenhouse;

[0117] S6, obtaining a control instruction applied to the internal environment parameters of the greenhouse according to the optimal control strategy;

[0118] S7, taking the time of sending the control instruction for the first time as the starting point, collecting the internal environment data of the greenhouse every preset time interval through the data acquisition device to obtain the post-environment data, and determining whether the post-environment data is within the normal range;

[0119] If yes, the post-environment data is within the normal range, and the present control is ended;

[0120] If no, return to step S4 for circulation.

[0121] Specifically, the time of sending the control instruction of the internal environment data of the greenhouse for the first time is t, and the preset time interval is Δt. After the interval time interval Δt from the first sending of the control instruction, the internal environment data of the greenhouse at t+Δt is collected through the data acquisition device.

[0122] The post-environment data collected at t+Δt is compared with the normal range of the preset greenhouse crop growth environment data. If the post-environment data collected at t+Δt is within the normal range, the present control is ended. If the post-environment data collected at t+Δt is not within the normal range, return to step S24 and initialize the initial population of the genetic algorithm with the post-environment data as the new current environment data, reacquire the corresponding control instruction and send it to the corresponding execution device. Every time interval Δt, the above operation is repeated until the internal environment of the greenhouse is adjusted to the normal range suitable for crop growth. By adjusting the control strategy in time according to the change of the internal environment of the greenhouse, the energy consumption can be greatly reduced, and the advantages of high intelligence and low management cost are achieved.

[0123] It should be noted that the two adjacent current environment data, the previous acquisition is the front current environment data, and the latter acquisition is the rear environment data.

[0124] The embodiment takes the time of sending the control instruction for the first time as the starting point, collects the rear environment data in the greenhouse every interval preset time period, and judges whether the rear environment data is in the normal range. If the rear environment data is not in the normal range, the control instruction is reacquired and sent to the corresponding execution device, so as to adjust the control strategy in time according to the change of the environment data in the greenhouse, until the environment data in the greenhouse is adjusted to the normal range, the energy waste can be effectively avoided, the advantages of high intelligence and low management cost are obtained, and the practicability is enhanced.

[0125] The control system of the environment in the greenhouse in the embodiment is the same as the corresponding control system in the foregoing embodiment, and has the beneficial effects of the corresponding system embodiment, which will not be described here.

[0126] The present application obtains the optimal control strategy of the environment control in the greenhouse by using the compensation fuzzy neural network based on the genetic algorithm optimization, considering the constraint relationship between the multiple environment variables in the greenhouse from the global and the sample rules based on the time change, solves the technical problems that the greenhouse environment cannot be accurately controlled, energy waste is easy to cause, and the management cost is high.

[0127] Embodiment three

[0128] The embodiment of the present application transmits the current environment data of the current time in the greenhouse collected by the data acquisition device to the field controller through the Internet of Things combined with the Internet technology. The field controller intelligently analyzes the current environment data. When the situation of the growth of crops is not conducive, the optimal control strategy is automatically sent to the main controller. The main controller generates the corresponding control instruction according to the optimal control strategy, sends the control instruction to the corresponding execution device through the field controller, and displays it on the man-machine interaction device. It is not only convenient to adjust the environment in the greenhouse in time, but also can warn the staff, and from many aspects, ensure that the crops in the greenhouse have a favorable growth environment, has the advantages of stable operation, fast and accurate and the like.

[0129] Please refer to Figure 4 The embodiment provides a system for realizing the above-mentioned control method based on big data application, which comprises a control module 100, a data acquisition device 200, a cloud storage platform 300 and an execution device 400.

[0130] The control module 100 comprises a main controller and a plurality of field controllers, one field controller is installed in each greenhouse, which is used to process the current environmental data collected by the data collection device 200 in the greenhouse to obtain an optimal control strategy, and send the current environmental data and the optimal control strategy to the main controller through the network; the main controller is installed in the control center, and is used to receive the current environmental data and the optimal control strategy sent by each field controller, and send a control instruction to the corresponding field controller according to the optimal control strategy.

[0131] The control module 100 further comprises a man-machine interaction device, the man-machine interaction device establishes data transmission with the main controller, and the man-machine interaction device is used for a user to view the internal environment data of the greenhouse and the control strategy taken, so as to facilitate the staff to deal with emergency situations.

[0132] The data collection device 200 is installed to establish data communication with the field controller through the Internet of Things, and the data collection device 200 is used to collect the current environmental data in the greenhouse. Specifically, the data collection device 200 comprises a temperature sensor, a humidity sensor, a carbon dioxide concentration sensor and a light intensity collector.

[0133] The temperature sensor is used to collect the temperature of the internal environment of the agricultural greenhouse, the humidity sensor is used to collect the humidity of the internal environment of the agricultural greenhouse, the carbon dioxide concentration sensor is used to collect the carbon dioxide concentration of the internal environment of the agricultural greenhouse, and the light intensity collector is used to collect the light intensity of the internal environment of the agricultural greenhouse.

[0134] The cloud storage platform 300 establishes data communication with the field main controller through a wireless network, and the cloud storage platform 300 is used to store historical environmental data for various greenhouse crops at each stage of the whole life cycle, which is screened from massive historical data by big data technology, and has a storage function, a query function, an internal environment evaluation function and a real-time early warning function.

[0135] The execution device 400 establishes data communication with the field controller through the Internet of Things, and the execution device is used to adjust the environment in the greenhouse. Specifically, the execution device 400 comprises a ventilation device, a roller shutter device, a water pump, a carbon dioxide generator and a light supplement lamp.

[0136] The ventilation device and the roller shutter device are used to adjust the temperature in the internal environment of the greenhouse according to the corresponding actions of the control instruction; the water pump is used to adjust the humidity in the internal environment of the greenhouse according to the corresponding actions of the control instruction; the carbon dioxide generator is used to adjust the carbon dioxide concentration in the internal environment of the greenhouse according to the corresponding actions of the control instruction; and the light supplement lamp is used to adjust the light intensity in the internal environment of the greenhouse according to the corresponding actions of the control instruction.

[0137] Through the embodiment, the current environment data of the current moment inside the agricultural greenhouse collected by the data collection equipment is transmitted to the field controller through the Internet of Things combined with Internet technology, the field controller intelligently analyzes the current environment data, when the situation is not conducive to the growth of crops, the optimal control strategy is automatically sent to the main controller, then the main controller generates corresponding control instructions according to the optimal control strategy, the control instructions are sent to the corresponding execution equipment through the field controller, and the control instructions are displayed on the man-machine interaction equipment, which not only facilitates the timely adjustment of the internal environment of the agricultural greenhouse, but also can warn the staff, and from many aspects, ensures that the crops in the agricultural greenhouse have a favorable growth environment, and has the advantages of stable operation, rapidness and accuracy.

[0138] Each of the embodiments in the specification is described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same and similar parts between the embodiments can be referred to each other. For the above embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the related parts can refer to the part of the method embodiment.

[0139] Those skilled in the art can understand that all or part of the steps of the above-mentioned embodiment methods can be completed by a program instructing related hardware, therefore, the present application can adopt a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer usable program codes.

[0140] The above embodiments are described in detail, the principles and implementation manners of the present application are described by applying specific examples, the above embodiment is only used to help understand the method of the present application and its core idea; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation manner and application range will be changed, and the above is not used to limit the present application.

Claims

1. A method of regulating the internal environment of a greenhouse for the cultivation of crops, characterized in that, The regulation method comprises the following steps: S1, obtaining current environment data of an agricultural greenhouse at a current time collected by a data acquisition device; S2, inputting the current environment data into an environment prediction model to obtain an evolution rule of the greenhouse internal environment data based on time; S3, constructing a fitness function of a genetic algorithm according to the evolution rule of the greenhouse internal environment data based on time, comprising: Setting represents the inside environment temperature of the greenhouse at the current time, represents the inside environment temperature of the greenhouse predicted by the environment prediction model to the next monitoring period, and the expression of the fitness function is: If , it indicates that the air temperature in the greenhouse interior environment gradually increases from the current time to the next monitoring period. The expression of the fitness function for the regulation of the greenhouse interior environment in this process is: ； If , it indicates that the air temperature of the greenhouse interior environment gradually decreases from the current time to the next monitoring period. The expression of the fitness function for the management of the greenhouse interior environment in this process is: ； In the above formula, denotes the average value of the variable , denotes the number of output variables of the environmental prediction model , denotes the maximum value of the variable , has a value range of [0, 1]; S4, initializing the initial population number of the genetic algorithm according to the current environment data, and encoding the initial population, the expression being: ； In the above formula, a control strategy for a greenhouse environment, respectively represent the control amount and control duration of the environmental parameters, , represent the type of environmental parameters contained in the current environmental data; S5, repeatedly selecting, crossing and mutating the initial population according to the fitness function to solve the optimal regulation strategy of the greenhouse internal environment; S6, obtaining a regulation instruction applied to the greenhouse internal environment parameter according to the optimal regulation strategy; S7, sending the regulation instruction to the corresponding execution device to regulate the greenhouse internal environment through the execution device.

2. The method of claim 1, wherein, In step S2, the current environment data is input into the environment prediction model to obtain the evolution rule of the greenhouse internal environment data based on time, and the specific process comprises the following steps: S201, inputting the current environment data into the environment prediction model to initialize the network parameters; S202, presetting the fuzzy segmentation number of the environment prediction model as 2, the learning rate as 1.0, the compensation degree as 0.8, and the expected error as 0.001, and taking the initialized parameters as the starting point to perform the compensation fuzzy reasoning and error back propagation process, and iteratively until the predetermined expected error is reached; S203, the environment prediction model after the circulation iteration determines the environment data of the inside of the agricultural greenhouse to the next monitoring period according to the current environment data , the environment data is the time-based evolution rule of the inside environment data of the agricultural greenhouse, and the expression of the environment data is: ； In the above formula, an output variable of the environment prediction model, a connection weight between the input layer and the output layer, an input variable of the environment prediction model, a membership function of each input variable, the number of nodes of the input layer.

3. The method of claim 1, wherein the step of modulating comprises: In step S5, the method for solving the optimal regulation strategy of the greenhouse internal environment is as follows: S51, solving the objective function of the regulation strategy corresponding to each individual in the population; The objective function expression is: ； In the above formula, represents the total amount of energy consumed by the greenhouse internal environment regulation, represents the type of environmental parameters contained in the current environmental data, represents the average amount of energy consumption of the execution device adjusting the first environmental parameter in the first time period, represents the length of the greenhouse environment regulation period; S52, determining the fitness value according to the fast non-dominated sorting and crowding degree calculation of the initial population according to the objective function; S53, repeatedly selecting, crossing and mutating the initial population according to the fitness function, evaluating and deciding from a global perspective, and outputting the optimal regulation strategy of the greenhouse environment.

4. The method of claim 1, wherein the step of modulating comprises: In step S7, the regulation instruction is sent to the corresponding execution device to regulate the greenhouse internal environment through the execution device, which can also be set as: Taking the time of sending the regulation instruction for the first time as the starting point, the data acquisition device collects the greenhouse internal environment data every interval of a preset time period to obtain the post-environment data, and judges whether the post-environment data is within the normal range; If yes, the post-environment data is within the normal range, and the regulation is ended; If not, return to step S4 for circulation.

5. A system for implementing the method of regulating according to any of the preceding claims 1-4, characterized in that, The system comprises a control module (100), a data acquisition device (200), a cloud storage platform (300) and an execution device (400). The control module (100) comprises a main controller and a plurality of field controllers; the field controllers are used to process current environmental data collected by data collection devices (200) in the greenhouse to obtain optimal regulation strategies, and send the current environmental data and the optimal regulation strategies to the main controller through a network; the main controller is used to receive the current environmental data and the optimal regulation strategies sent by each field controller, and send regulation instructions to the corresponding field controllers according to the optimal regulation strategies; The data collection devices (200) establish data communication with the field controllers, and the data collection devices (200) are used to collect current environmental data inside the greenhouse; The cloud storage platform (300) establishes data communication with the field controllers, and the cloud storage platform (300) is used to store greenhouse environmental data of various greenhouse crops at various stages in the whole life cycle; The execution devices (400) establish data communication with the field controllers, and the execution devices (400) are used to adjust the environment in the greenhouse.

6. The system of claim 5, wherein, The control module (100) establishes data transmission with a human-computer interaction device, and the human-computer interaction device is used for users to view the greenhouse environment and perform control operations.

7. The system of claim 5, wherein, The execution devices (400) comprise ventilation devices, roller shutter devices, water pumps, carbon dioxide generators and light supplement lamps; The ventilation devices and the roller shutter devices are used to execute corresponding actions to adjust the temperature inside the greenhouse according to the regulation instructions; The water pumps are used to execute corresponding actions to adjust the humidity inside the greenhouse according to the regulation instructions; The carbon dioxide generators are used to execute corresponding actions to adjust the carbon dioxide concentration inside the greenhouse according to the regulation instructions; The light supplement lamps are used to execute corresponding actions to adjust the light intensity inside the greenhouse according to the regulation instructions.

Images