Greenhouse temperature field local prediction method based on differential steganography modeling

Abstract

The invention discloses a greenhouse temperature field local prediction method based on differential steganography modeling. The method comprises the following steps: S1, collecting or simulating limited temperature acquisition point data in a greenhouse; s2, preprocessing the collected data of the temperature acquisition points; s3, applying a differential steganography modeling technology on a macroscopic scale, performing differential processing on the preprocessed temperature acquisition point data, and establishing a global temperature field model; s4, on the scale of the greenhouse area, differential processing is carried out on the temperature data in the area by applying a differential steganography modeling technology, and an area-level temperature field model is established; s5, performing differential processing on the temperature data of the key position by applying a differential steganography modeling technology on the micro viewpoint scale, and establishing a micro-level temperature field model; s6, combining the models in different scales from S3 to S5 to generate global and local temperature field prediction data in the greenhouse; and S7, adjusting a temperature control system in the greenhouse based on the global and local temperature field prediction data. According to the invention, accurate monitoring and intelligent adjustment of the internal temperature of the greenhouse are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of greenhouse temperature field, and in particular to a local prediction method of greenhouse temperature field based on differential steganography modeling. Background Art

[0002] In the field of environmental control in agricultural greenhouses, the prior art usually relies on a large number of sensors for real-time monitoring of environmental variables, especially temperature monitoring, which directly affects the growth of crops. However, to accurately monitor the temperature in various areas and corners of the greenhouse, it is usually necessary to arrange a large number of temperature sensors at different locations. Although this method can improve the coverage and accuracy of temperature monitoring, it also faces several challenges. First, the large-scale deployment of sensors means an increase in hardware costs, especially in large greenhouses. The procurement, installation and maintenance costs of sensors are high. In addition, the deployment process of sensors is complicated, especially in some areas where it is difficult to install sensors, which increases the difficulty of implementation. Secondly, the data collection frequency and coverage of sensors are limited. The temperature data collection in some areas may be insufficient, resulting in inaccuracy in global temperature monitoring.

[0003] On the other hand, the existing temperature control system is usually adjusted based on the global temperature average, and lacks precise control of different areas or micro details in the greenhouse. This control method is ineffective when faced with uneven temperature distribution in the greenhouse, especially when different crops in the greenhouse have different temperature requirements. It is impossible to achieve fine adjustment, which not only leads to energy waste, but may also affect the growth quality and yield of crops. In addition, the existing technology lacks flexibility in dealing with local temperature fluctuations inside the greenhouse, and cannot respond quickly to local areas, resulting in delayed adjustment of the temperature control system.

[0004] To sum up, the existing technology has the following main shortcomings: first, the temperature monitoring inside the greenhouse relies on a large number of sensors, resulting in excessively high hardware costs and complex sensor deployment; second, the existing temperature control system lacks precise control of the temperature of the local area and cannot effectively deal with the uneven temperature distribution in the greenhouse; finally, the temperature control reaction speed of the existing technology is slow and cannot respond to local temperature changes in time, affecting the efficiency of the temperature control system and the growth quality of crops. Summary of the invention

[0005] One purpose of the present invention is to propose a local prediction method for greenhouse temperature field based on differential steganographic modeling, which realizes accurate monitoring and intelligent regulation of the internal temperature of the greenhouse.

[0006] A method for local prediction of greenhouse temperature field based on differential steganographic modeling according to an embodiment of the present invention comprises the following steps:

[0007] S1. Collect or simulate limited temperature collection point data inside the greenhouse;

[0008] S2. Preprocessing the collected temperature collection point data;

[0009] S3. Apply differential steganography modeling technology on a macro scale, perform differential processing on the pre-processed temperature collection point data, and establish a global temperature field model;

[0010] S4. At the greenhouse regional scale, differential steganography modeling technology is used to perform differential processing on the temperature data in the region and establish a regional temperature field model;

[0011] S5. At the microscopic scale, differential steganography modeling technology is used to perform differential processing on the temperature data of key positions and establish a microscopic temperature field model;

[0012] S6. Combine the models at different scales in S3-S5 to generate global and local temperature field prediction data inside the greenhouse;

[0013] S7. Adjust the temperature control system inside the greenhouse based on the global and local temperature field prediction data.

[0014] Optionally, the S1 includes the following steps:

[0015] S11. Determine the location of the temperature collection point inside the greenhouse, set the spatial range of the greenhouse to Ω, P(x, y, z) to be a collection point in the greenhouse, and the measurement value of the temperature sensor at the collection point to be T P (t), where t is time. Combined with the number of sensors N, the finite temperature data of all collection points can be expressed as:

[0016]

[0017] Among them, f(x i ,y i ,z i ,t) is the temperature distribution function of space and time, ∈ i Measuring noise for sensors;

[0018] S12. Generate limited temperature collection point data at locations where sensors are not installed, build a multivariate temperature prediction model, and generate simulated limited temperature collection point data T based on sensor data and environmental variables. sim (P):

[0019]

[0020] Among them, α and β are the adjustment parameters of the simulation model. is the second-order derivative of the temperature gradient, which is used to simulate the diffusion behavior of the temperature field. G(x, y, z) is the external disturbance function in the greenhouse environment, including humidity and light.

[0021] S13. Use the weighted average method to calculate the actual measured finite temperature collection point data With the simulated data T sim (P, t) is fused:

[0022]

[0023] Among them, w1 and w2 are fusion weights;

[0024] S14. Based on the error estimation method, the fused data T 融合 (P,t) Perform quality inspection and verification:

[0025]

[0026] Among them, γ is the correction coefficient, is the expected temperature value, which is used to judge the degree of deviation of the data;

[0027] S15. The corrected finite temperature acquisition point data T 修正 (P, t) is stored in the database in the form of a space-time matrix:

[0028]

[0029] Where T represents the stored temperature field matrix, N is the number of sensors, and k is the total number of time points.

[0030] Optionally, S3 includes the following steps:

[0031] S31. The pre-processed and corrected temperature collection point data T 修正 (P, t) is input into the macro-scale temperature field modeling framework;

[0032] S32. Apply differential steganography technology to the temperature collection point data T 修正 (P, t) is differentially processed to calculate the global temperature field gradient

[0033]

[0034] in, It represents the gradient change of the temperature field at space P, capturing the trend of temperature change with spatial coordinates;

[0035] S33. Based on the difference processing results, calculate the global temperature field diffusion model inside the greenhouse:

[0036]

[0037] in, is the second-order derivative of the temperature field, indicating the diffusion rate of temperature, α1 is the temperature diffusion coefficient, β1 is the external factor influence coefficient, and F(P,t) is the external environmental factor function, including the influence of light, humidity and ventilation on the temperature field;

[0038] S34. Using steganographic modeling technology to expand the details of the global temperature field diffusion model to generate the greenhouse global temperature field model T 全局 (P, t), the greenhouse global temperature field model output is the temperature distribution of each collection point in the greenhouse and its surrounding areas:

[0040]

[0041] Among them, T 全局 (P, t) represents the predicted value of the global temperature field at spatial position P (x, y, z) and time t, is the Laplace operator of the temperature field, which indicates the degree of temperature diffusion in space. λ1, λ2, λ3 and λ4 are weights. G(x, y, z) is the spatial function of greenhouse environmental factors. H(t) is the function of environmental factors that change over time. Represents the time derivative of environmental factors, which is used to describe the immediate impact of environmental changes on the temperature field. represents the rate of change of the corrected temperature field over time, α2 is the coefficient that controls temperature diffusion, The corrected temperature data is subjected to Fourier transform to extract the frequency components in the temperature field.

[0042] Optionally, S4 includes the following steps:

[0043] S41. Determine the various areas inside the greenhouse k , the greenhouse is divided into several areas with different temperature requirements, and the temperature collection point position in each area is P k (x k ,y k ,z k ), where k represents the kth region, P k Indicates the temperature collection points in the area;

[0044] S42. Collect temperature data in each area Perform differential processing to calculate the gradient change of the regional temperature field

[0045]

[0046] in, Represents the temperature field in region k at the spatial coordinate P k (x k ,y k ,z k ) to capture the temperature variation characteristics in the region with space;

[0047] S43. According to the gradient change A regional temperature diffusion model is established, which is expressed as:

[0048]

[0049] Among them, α k is the temperature diffusion coefficient of region k, is the second-order derivative of the temperature field in the region, β k is the external factor influence coefficient of region k, F k (P k ,t) is the function of the external environmental factors of the region;

[0050] S44. Combined with the temperature distribution characteristics within the region, the regional temperature field is expanded through steganographic modeling technology to generate a regional temperature field model T 区域 (P k ,t):

[0051]

[0052] Among them, λ k2 and λ k1 is the weight coefficient in the regional temperature field model, G k (x k ,y k ,z k ) and H k (t) represent the spatial and temporal environmental impact functions of region k respectively.

[0053] Optionally, S5 includes the following steps:

[0054] S51. Determine the key position inside the greenhouse and set the temperature collection point position at the micro scale as P m 9x m ,y m ,z m ), where m represents the identifier of the key position, P m Indicates the temperature collection point at the microscopic scale;

[0055] S52. Temperature data for key locations Perform differential processing to calculate the temperature field gradient at the microscopic scale

[0056]

[0057] in, It represents the gradient change of the temperature field at key positions at the microscopic scale, and is used to capture the temperature distribution of local details;

[0058] S53. Based on gradient changes The temperature diffusion model at the microscopic scale is established and expressed as:

[0059]

[0060] Among them, α m is the temperature diffusion coefficient at the microscopic scale, is the second-order derivative of the temperature field, β m is the external factor influence coefficient, F m (P m ,t) is the function of external environmental factors at key locations;

[0061] S54. Combine the temperature variation characteristics of key positions and use steganographic modeling technology to model the temperature field at the microscopic scale to generate a microscopic temperature field model T 微观 (P m ,t):

[0062]

[0063] Among them, λ m1 and λ m2 is the weight coefficient of the micro-scale model, G m (x m ,y m ,z m ) and H m (t) represent the spatial and temporal environmental impact functions of key locations respectively.

[0064] Optionally, the S6 includes the following steps:

[0065] S62. Fusion processing is performed on the global temperature field model, the regional temperature field model and the microscopic temperature field model to construct a comprehensive temperature field prediction model and set the weight parameters ω of the models at different scales. 全局 ,ω 区域 ,ω 微观 , generate the fused temperature field prediction data T 融合 (P,t):

[0067]

[0068] Among them, ω 全局 ,ω区域 ,ω 微观 is the weight of models of different scales, indicating the contribution of each model to the overall temperature field prediction; is the second-order spatial derivative of the global temperature field to capture the global temperature diffusion; α 全局 is the parameter that controls the decay of the global temperature field over time; is the second-order derivative of the regional temperature field, combined with the regional environmental factor function G k (x k ,y k ,z k ) and time function H k The time derivative of (t); λ m is the adjustment coefficient of the micro-scale temperature field, is the second-order time derivative of the microscopic temperature field, δ m Control the time-varying decay of the microscopic temperature field, Ω, Ω k ,Ω m denote the spatial extents at the global, regional, and microscopic scales, respectively;

[0069] S63. Comprehensive temperature field prediction data T 融合 (P, t) performs error analysis and correction by calculating the error ∈ 全局 ,∈ 区域 ,∈ 微观 The corrected temperature field prediction data is:

[0070]

[0071] Among them, ∈ 全局 ,∈ 区域 ,∈ 微观 are the error terms of the global, regional, and microscale models, respectively;

[0072] S64. The corrected temperature field prediction data T 修正 (P, t) is stored as the final prediction data of the global and local temperature fields inside the greenhouse. The generated prediction data is represented by:

[0073] T 最终 =[T 修正 (P1,t1),T 修正 (P2,t2),,T 修正 (P N ,t k )];

[0074] Among them, T 最终 Represents the matrix form of the final prediction data, P i is the location of the collection point inside the greenhouse, t k is each time point, and N is the number of sensors.

[0075] Optionally, the S7 includes the following steps:

[0076] S71. Based on the corrected global and local temperature field prediction data T 修正 (P, t) Determine the temperature control requirements of each area inside the greenhouse, build the adjustment strategy of the temperature control system, and set the target temperature of the temperature control system as T 目标 (P,t);

[0077] S72. Calculate the current actual temperature T in each area 当前 (P,t) and target temperature T 目标 The difference between (P, t) generates the temperature control adjustment instruction:

[0078] ΔT 调节 (P,t)=T 目标 (P,t)-T 当前 (P,t);

[0079] Where, ΔT 调节 (P,t) is the temperature adjustment required at each collection point, a positive value indicates that the temperature needs to be increased, and a negative value indicates that the temperature needs to be decreased;

[0080] S73. Based on the temperature difference ΔT in different areas 调节 (P, t) By adjusting the heating or cooling equipment of the greenhouse temperature control system, the temperature of each area can be adjusted in real time;

[0081] S74. Real-time monitoring of the adjustment effect of the temperature control system to obtain the adjusted temperature T 调节后 (P,t), and the target temperature T 目标 (P, t) for comparison, if the error ∈ 调节 If (P,t) exceeds the preset range, the temperature control strategy is automatically adjusted to correct the adjustment rate of the temperature control device;

[0082] S75. Store the adjustment data of the temperature control system into the database, and record the temperature changes in each area, equipment adjustment parameters and adjustment effects.

[0083] The beneficial effects of the present invention are:

[0084] (1) The present invention introduces differential steganography modeling technology, utilizes limited temperature collection point data, and establishes a global temperature field model through differential processing and steganography technology. The spatial gradient and diffusion rate of the temperature field are calculated at the macro, regional and micro scales, respectively, so as to accurately predict the temperature distribution inside the greenhouse. By integrating the global, regional and local temperature field models, the dependence on a large number of sensors is effectively reduced. Under the condition of limited sensor data, a complete temperature field distribution can still be constructed, which significantly reduces the hardware cost and deployment difficulty, and improves the coverage and accuracy of greenhouse environmental monitoring.

[0085] (2) The present invention applies differential steganographic modeling technology at the regional and micro scales to capture the temperature change characteristics of different areas and key locations inside the greenhouse. By accurately modeling the temperature field of each area, the system can perform local temperature control according to the actual needs of different crops, avoiding the deficiency of traditional global temperature control methods that cannot effectively respond to local temperature needs. The present invention provides real-time monitoring and dynamic adjustment functions, which can quickly respond to local temperature changes and accurately adjust the output of local temperature control equipment, effectively improving energy utilization efficiency and consistency of crop growth.

[0086] (3) The present invention constructs a comprehensive temperature field prediction model by fusing the temperature field models at three different scales: global, regional and microscopic. The fusion model is then analyzed and corrected for errors. The system calculates the gradient change and diffusion behavior of the temperature field at different scales. By introducing multi-level adjustment parameters, the prediction results of the temperature field are ensured to be more accurate. Combined with the error correction mechanism, the strategy of the temperature control system can be adaptively adjusted to ensure the real-time and accuracy of temperature regulation. This further optimizes the monitoring and control of the temperature field inside the greenhouse and improves the intelligence level of the temperature control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0088] Figure 1 This is a flow chart of a method for local prediction of greenhouse temperature field based on differential steganographic modeling proposed by the present invention. DETAILED DESCRIPTION

[0089] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0090] refer to Figure 1 , a greenhouse temperature field local prediction method based on differential steganography modeling, comprising the following steps:

[0091] S1. Collect or simulate limited temperature collection point data inside the greenhouse;

[0092] S2. Preprocessing the collected temperature collection point data;

[0093] S3. Apply differential steganography modeling technology on a macro scale, perform differential processing on the pre-processed temperature collection point data, and establish a global temperature field model;

[0094] S4. At the greenhouse regional scale, differential steganography modeling technology is used to perform differential processing on the temperature data in the region and establish a regional temperature field model;

[0095] S5. At the microscopic scale, differential steganography modeling technology is used to perform differential processing on the temperature data of key positions and establish a microscopic temperature field model;

[0096] S6. Combine the models at different scales in S3-S5 to generate global and local temperature field prediction data inside the greenhouse;

[0097] S7. Adjust the temperature control system inside the greenhouse based on the global and local temperature field prediction data.

[0098] In this implementation, S1 includes the following steps:

[0099] S11. Determine the location of the temperature collection point inside the greenhouse, set the spatial range of the greenhouse to Ω, P(x, y, z) to be a collection point in the greenhouse, and the measurement value of the temperature sensor at the collection point to be T P (t), where t is time. Combined with the number of sensors N, the finite temperature data of all collection points can be expressed as:

[0100]

[0101] Among them, f(x i ,y i ,z i ,t) is the temperature distribution function of space and time, ∈ i Measuring noise for sensors;

[0102] S12. Generate limited temperature collection point data at locations where sensors are not installed, build a multivariate temperature prediction model, and generate simulated limited temperature collection point data T based on sensor data and environmental variables. sim (P):

[0103]

[0104] Among them, α and β are the adjustment parameters of the simulation model. is the second-order derivative of the temperature gradient, which is used to simulate the diffusion behavior of the temperature field. G(x, y, z) is the external disturbance function in the greenhouse environment, including humidity and light.

[0105] S13. Use the weighted average method to calculate the actual measured finite temperature collection point data With the simulated data T sim (P, t) is fused:

[0106]

[0107] Among them, w1 and w2 are fusion weights;

[0108] S14. Based on the error estimation method, the fused data T 融合 (P,t) Perform quality inspection and verification:

[0109]

[0110] Among them, γ is the correction coefficient, is the expected temperature value, which is used to judge the degree of deviation of the data;

[0111] S15. The corrected finite temperature acquisition point data T 修正 (P, t) is stored in the database in the form of a space-time matrix:

[0112]

[0113] Where T represents the stored temperature field matrix, N is the number of sensors, and k is the total number of time points.

[0114] In this implementation, S3 includes the following steps:

[0115] S31. The pre-processed and corrected temperature collection point data T 修正 (P, t) is input into the macro-scale temperature field modeling framework;

[0116] S32. Apply differential steganography technology to the temperature collection point data T 修正 (P, t) is differentially processed to calculate the global temperature field gradient

[0117]

[0118] in, It represents the gradient change of the temperature field at space P, capturing the trend of temperature change with spatial coordinates;

[0119] S33. Based on the difference processing results, calculate the global temperature field diffusion model inside the greenhouse:

[0120]

[0121] in, is the second-order derivative of the temperature field, indicating the diffusion rate of temperature, α1 is the temperature diffusion coefficient, β1 is the external factor influence coefficient, and F(P,t) is the external environmental factor function, including the influence of light, humidity and ventilation on the temperature field;

[0122] S34. Using steganographic modeling technology to expand the details of the global temperature field diffusion model to generate the greenhouse global temperature field model T全局 (P, t), the greenhouse global temperature field model output is the temperature distribution of each collection point in the greenhouse and its surrounding areas:

[0123]

[0124] Among them, T 全局 (P, t) represents the predicted value of the global temperature field at spatial position P (x, y, z) and time t, is the Laplace operator of the temperature field, which indicates the degree of temperature diffusion in space. λ1, λ2, λ3 and λ4 are weights. G(x, y, z) is the spatial function of greenhouse environmental factors. H(t) is the function of environmental factors that change over time. Represents the time derivative of environmental factors, which is used to describe the immediate impact of environmental changes on the temperature field. represents the rate of change of the corrected temperature field over time, α2 is the coefficient that controls temperature diffusion, The corrected temperature data is subjected to Fourier transform to extract the frequency components in the temperature field.

[0125] In this implementation, S4 includes the following steps:

[0126] S41. Determine the various areas inside the greenhouse k , the greenhouse is divided into several areas with different temperature requirements, and the temperature collection point position in each area is P k (x k ,y k ,z k ), where k represents the kth region, P k Indicates the temperature collection points in the area;

[0127] S42. Collect temperature data in each area Perform differential processing to calculate the gradient change of the regional temperature field

[0128]

[0129] in, Represents the temperature field in region k at the spatial coordinate P k (x k ,y k ,z k ) to capture the temperature variation characteristics in the region with space;

[0130] S43. According to the gradient change A regional temperature diffusion model is established, which is expressed as:

[0131]

[0132] Among them, α k is the temperature diffusion coefficient of region k, is the second-order derivative of the temperature field in the region, β k is the external factor influence coefficient of region k, F k (P k ,t) is the function of the external environmental factors of the region;

[0133] S44. Combined with the temperature distribution characteristics within the region, the regional temperature field is expanded through steganographic modeling technology to generate a regional temperature field model T 区域 (P k ,t):

[0134]

[0135] Among them, λ k2 and λ k1 is the weight coefficient in the regional temperature field model, G k (x k ,y k ,z k ) and H k (t) represent the spatial and temporal environmental impact functions of region k respectively.

[0136] In this implementation, S5 includes the following steps:

[0137] S51. Determine the key position inside the greenhouse and set the temperature collection point position at the micro scale as P m (x m ,y m ,z m ), where m represents the identifier of the key position, P m Indicates the temperature collection point at the microscopic scale;

[0138] S52. Temperature data for key locations Perform differential processing to calculate the temperature field gradient at the microscopic scale

[0139]

[0140] in, It represents the gradient change of the temperature field at key positions at the microscopic scale, and is used to capture the temperature distribution of local details;

[0141] S53. Based on gradient changes The temperature diffusion model at the microscopic scale is established and expressed as:

[0142]

[0143] Among them, α m is the temperature diffusion coefficient at the microscopic scale, is the second-order derivative of the temperature field, β m is the external factor influence coefficient, F m (P m ,t) is the function of external environmental factors at key locations;

[0144] S54. Combine the temperature variation characteristics of key positions and use steganographic modeling technology to model the temperature field at the microscopic scale to generate a microscopic temperature field model T 微观 (P m ,t):

[0145]

[0146] Among them, λ m1 and λ m2 is the weight coefficient of the micro-scale model, G m (x m ,y m ,z m ) and H m (t) represent the spatial and temporal environmental impact functions of key locations respectively.

[0147] In this implementation, S6 includes the following steps:

[0148] S62. Fusion processing is performed on the global temperature field model, the regional temperature field model and the microscopic temperature field model to construct a comprehensive temperature field prediction model and set the weight parameters ω of the models at different scales. 全局 ,ω 区域 ,ω 微观 , generate the fused temperature field prediction data T 融合 (P,t):

[0149]

[0150] Among them, ω 全局 ,ω 区域 ,ω 微观 is the weight of models of different scales, indicating the contribution of each model to the overall temperature field prediction; is the second-order spatial derivative of the global temperature field to capture the global temperature diffusion; α 全局 is the parameter that controls the decay of the global temperature field over time; is the second-order derivative of the regional temperature field, combined with the regional environmental factor function G k (x k ,y k ,z k ) and time function H k The time derivative of (t); λm is the adjustment coefficient of the micro-scale temperature field, is the second-order time derivative of the microscopic temperature field, δ m Control the time-varying decay of the microscopic temperature field, Ω, Ω k ,Ω m denote the spatial extents at the global, regional, and microscopic scales, respectively;

[0151] S63. Comprehensive temperature field prediction data T 融合 (P, t) performs error analysis and correction by calculating the error ∈ 全局 ,∈ 区域 ,∈ 微观 The corrected temperature field prediction data is:

[0152]

[0153] Among them, ∈ 全局 ,∈ 区域 ,∈ 微观 are the error terms of the global, regional, and microscale models, respectively;

[0154] S64. The corrected temperature field prediction data T 修正 (P, t) is stored as the final prediction data of the global and local temperature fields inside the greenhouse. The generated prediction data is represented by:

[0155] T 最终 =[T 修正 (P1,t1),T 修正 (P2,t2),,T 修正 (P N ,t k )];

[0156] Among them, T 最终 Represents the matrix form of the final prediction data, P i is the location of the collection point inside the greenhouse, t k is each time point, and N is the number of sensors.

[0157] In this implementation, S7 includes the following steps:

[0158] S71. Based on the corrected global and local temperature field prediction data T 修正 (P, t) Determine the temperature control requirements of each area inside the greenhouse, build the adjustment strategy of the temperature control system, and set the target temperature of the temperature control system as T 目标 (P,t);

[0159] S72. Calculate the current actual temperature T in each area 当前 (P,t) and target temperature T 目标The difference between (P, t) generates the temperature control adjustment instruction:

[0160] ΔT 调节 (P,t)=T 目标 (P,t)-T 当前 (P,t);

[0161] Where, ΔT 调节 (P,t) is the temperature adjustment required at each collection point, a positive value indicates that the temperature needs to be increased, and a negative value indicates that the temperature needs to be decreased;

[0162] S73. Based on the temperature difference ΔT in different areas 调节 (P, t) By adjusting the heating or cooling equipment of the greenhouse temperature control system, the temperature of each area can be adjusted in real time;

[0163] S74. Real-time monitoring of the adjustment effect of the temperature control system to obtain the adjusted temperature T 调节后 (P,t), and the target temperature T 目标 (P, t) for comparison, if the error ∈ 调节 If (P,t) exceeds the preset range, the temperature control strategy is automatically adjusted to correct the adjustment rate of the temperature control device;

[0164] S75. Store the adjustment data of the temperature control system into the database, and record the temperature changes in each area, equipment adjustment parameters and adjustment effects.

[0165] Embodiment 1:

[0166] From November 2023 to January 2024, we conducted tests in a smart greenhouse in a modern agricultural park in City A. The greenhouse covers an area of ​​about 4,000 square meters and mainly grows high-value-added crops such as cucumbers, tomatoes and strawberries. In order to verify the feasibility and effectiveness of the local prediction method of greenhouse temperature field based on differential steganographic modeling of the present invention, we upgraded the temperature control system in the greenhouse and compared the traditional method with the prediction method of the present invention.

[0167] At 2 p.m. on December 15, 2023, the temperature sensor data in the strawberry area of ​​the greenhouse suddenly showed abnormal fluctuations. The temperature fluctuated by more than 4°C, rising instantly from the original 19°C to 24°C, and maintained at a high level. Through the temperature field prediction system of the present invention, we immediately discovered the temperature anomaly in the area and quickly analyzed the cause.

[0168] 30 temperature sensors were deployed in the greenhouse, covering all major areas of the greenhouse, including the strawberry area, cucumber area and tomato area. The temperature collection frequency was set to record data every 15 minutes. At 1:45 pm on December 15, the system monitored that the temperature sensor data in the strawberry area showed a rapid rise in temperature from 19°C to 21°C. The system then immediately used differential steganography modeling technology to make global and local predictions of the temperature field in the area. The results showed that the temperature in the strawberry area will continue to rise in the next 30 minutes, which may cause heat damage to crops.

[0169] The prediction model is based on 30 sensor data collected in real time, combined with the greenhouse's regional temperature control historical data and external meteorological data (the external temperature was 3°C and the humidity was 70%). The global, regional and micro-scale prediction models generate a temperature field change curve for the strawberry area in the next 30 minutes. The system predicts that at around 2:30, the temperature in the strawberry area may exceed 25°C, which may cause the flowers and fruits of strawberry plants in the flowering period to fall and affect the yield.

[0170] At 2 o'clock, the system issued an early warning to the greenhouse manager based on the prediction results, indicating that the temperature in the strawberry area will continue to rise in the next 30 minutes, and suggested taking ventilation measures to reduce the local temperature. After receiving the early warning signal, the greenhouse manager immediately started the local ventilation equipment in the strawberry area, and at the same time lowered the output power of the heating equipment in the strawberry area through the temperature control system of the present invention.

[0171] At 2:15, the system automatically generated a real-time report, recording in detail the temperature change curves of each sensor in the past 30 minutes, and focusing on analyzing the temperature fluctuations in the strawberry area. The report showed that the temperature in the strawberry area reached 24.5°C at 2:05. Then, through the adjustment of the temperature control system and the activation of the ventilation equipment, the temperature gradually dropped and returned to around 21°C at 2:25, avoiding the risk of heat damage to strawberries.

[0172] We compared the effects of the traditional global temperature control method and the method of the present invention. Under the same environmental conditions, the traditional global temperature control system does not have an early warning mechanism for local areas and can only be regulated based on the overall average temperature of the greenhouse. The following is a specific comparison data of the two methods:

[0173] Temperature control accuracy comparison:

[0174] Traditional method: The temperature fluctuation range in the strawberry area is ±4℃, with the highest reaching 26℃ and the lowest being 18℃.

[0175] The method of the present invention: the temperature fluctuation range of the strawberry area is ±1.5°C, the highest is 24.5°C, and the lowest is 19.5°C, which successfully avoids the adverse effect of excessively high temperature on the strawberries.

[0176] Crop yield comparison:

[0177] Traditional method: At the end of January 2024, the average strawberry yield is 450 kg / mu and the cucumber yield is 500 kg / mu.

[0178] The method of the present invention increases the average yield of strawberries to 510 kg / mu and that of cucumbers to 540 kg / mu, and increases the yield of strawberries by 13.3% and that of cucumbers by 8%.

[0179] Energy consumption comparison:

[0180] Traditional method: Average daily energy consumption is 40 kWh.

[0181] The method of the invention has an average daily energy consumption of 32 kWh, which saves 20% of energy consumption.

[0182] At 1:45 p.m. on December 15, the temperature recorded by the sensor in the strawberry area was 19°C. The greenhouse management system concluded through the global prediction model of the method of the present invention that the temperature in the area would rise rapidly in the next 15 minutes, and it was expected that the temperature would reach above 24°C within 30 minutes. At this time, through the local temperature field model, the system found that the temperature fluctuations in the strawberry area were more drastic because the vents in the southeast direction were closed, resulting in local air stagnation.

[0183] The system further analyzed the temperature difference between the upper and lower parts of the strawberry plants through a micro-scale temperature field model and found that the temperature of the upper part of the plant was 23°C, while the temperature of the lower part was 21°C. The temperature difference may cause physiological disorders in the strawberry plants and affect fruit growth. The system automatically generated control instructions, suggesting opening the southeast vents and adjusting the power of the heater.

[0184] At 2:30, the system provided real-time feedback on the temperature control effect. The temperature in the strawberry area gradually dropped to 21°C, and the temperature difference problem was alleviated. Through this process, the present invention proves that in the case of limited sensor data, it is still possible to accurately predict and control the global and local temperature fields, and respond quickly to local abnormal conditions, thus avoiding potential damage to crops.

[0185] It can be seen from this embodiment that the local prediction method of greenhouse temperature field based on differential steganographic modeling of the present invention can still achieve accurate prediction and control of global and local temperature fields under the condition of limited sensor data inside the greenhouse. Through an effective temperature field prediction model, the system can monitor and predict temperature changes in real time and respond quickly, thus avoiding the problem of temperature damage to crops and significantly improving crop yield and energy utilization efficiency.

[0186] The present invention introduces differential steganography modeling technology, utilizes limited temperature collection point data, and establishes a global temperature field model through differential processing and steganography technology. The spatial gradient and diffusion rate of the temperature field are calculated at the macro, regional and micro scales, respectively, so as to accurately predict the temperature distribution inside the greenhouse. By fusing the global, regional and local temperature field models, the dependence on a large number of sensors is effectively reduced. Under the condition of limited sensor data, a complete temperature field distribution can still be constructed, which significantly reduces the hardware cost and deployment difficulty, and improves the coverage and accuracy of greenhouse environmental monitoring.

[0187] The present invention applies differential steganography modeling technology at regional and micro scales to capture the temperature change characteristics of different areas and key positions inside the greenhouse. Through accurate modeling of the temperature field in each area, the system can perform local temperature control according to the actual needs of different crops, avoiding the deficiency of traditional global temperature control methods that cannot effectively cope with local temperature needs. The present invention provides real-time monitoring and dynamic adjustment functions, which can quickly respond to local temperature changes and accurately adjust the output of local temperature control equipment, effectively improving energy utilization efficiency and consistency of crop growth.

[0188] The present invention constructs a comprehensive temperature field prediction model by fusing temperature field models at three different scales: global, regional and microscopic. The fusion model is then analyzed and corrected for errors. The system calculates the gradient change and diffusion behavior of the temperature field at different scales. By introducing multi-level adjustment parameters, the prediction results of the temperature field are more accurate. Combined with the error correction mechanism, the strategy of the temperature control system can be adaptively adjusted to ensure the real-time and accuracy of temperature regulation. This further optimizes the monitoring and control of the temperature field inside the greenhouse and improves the intelligence level of the temperature control system.

[0189] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A local prediction method for greenhouse temperature field based on differential steganography modeling, characterized in that: The steps include: S1. Collect or simulate limited temperature collection point data inside the greenhouse; S2. Preprocessing the collected temperature collection point data; S3. Apply differential steganography modeling technology on a macro scale, perform differential processing on the pre-processed temperature collection point data, and establish a global temperature field model; S4. At the greenhouse regional scale, differential steganography modeling technology is used to perform differential processing on the temperature data in the region and establish a regional temperature field model; S5. At the microscopic scale, differential steganography modeling technology is used to perform differential processing on the temperature data of key positions and establish a microscopic temperature field model; S6. Combine the models at different scales in S3-S5 to generate global and local temperature field prediction data inside the greenhouse; S7. Adjust the temperature control system inside the greenhouse based on the global and local temperature field prediction data.

2. According to claim 1, a method for local prediction of greenhouse temperature field based on differential steganographic modeling is characterized in that: The S1 comprises the following steps: S11. Determine the location of the temperature collection point inside the greenhouse, set the spatial range of the greenhouse to Ω, P(x, y, z) to be a collection point in the greenhouse, and the measurement value of the temperature sensor at the collection point to be T P (t), where t is time. Combined with the number of sensors N, the finite temperature data of all collection points can be expressed as: Among them, f(x i ,y i ,z i ,t) is the temperature distribution function of space and time, ∈ i Measuring noise for sensors; S12. Generate limited temperature collection point data at locations where sensors are not installed, build a multivariate temperature prediction model, and generate simulated limited temperature collection point data T based on sensor data and environmental variables. sim (P): Among them, α and β are the adjustment parameters of the simulation model. is the second-order derivative of the temperature gradient, which is used to simulate the diffusion behavior of the temperature field. G(x, y, z) is the external disturbance function in the greenhouse environment, including humidity and light. S13. Use the weighted average method to calculate the actual measured finite temperature collection point data With the simulated data T sim (P, t) is fused: Among them, w1 and w2 are fusion weights; S14. Based on the error estimation method, the fused data T 融合 (P,t) for quality inspection and verification: Among them, γ is the correction coefficient, is the expected temperature value, which is used to judge the degree of deviation of the data; S15. The corrected finite temperature acquisition point data T 修正 (P, t) is stored in the database in the form of a space-time matrix: Where T represents the stored temperature field matrix, N is the number of sensors, and k is the total number of time points.

3. According to claim 1, a method for local prediction of greenhouse temperature field based on differential steganographic modeling is characterized in that: The S3 comprises the following steps: S31. The pre-processed and corrected temperature collection point data T 修正 (P, t) is input into the macro-scale temperature field modeling framework; S32. Apply differential steganography technology to the temperature collection point data T 修正 (P, t) is differentially processed to calculate the global temperature field gradient in, It represents the gradient change of the temperature field at space P, capturing the trend of temperature change with spatial coordinates; S33. Calculate the global temperature field diffusion model inside the greenhouse based on the difference processing results: in, is the second-order derivative of the temperature field, indicating the diffusion rate of temperature, α1 is the temperature diffusion coefficient, β1 is the external factor influence coefficient, and F(P,t) is the external environmental factor function, including the influence of light, humidity and ventilation on the temperature field; S34. Using steganographic modeling technology to expand the details of the global temperature field diffusion model to generate the greenhouse global temperature field model T 全局 (P, t), the greenhouse global temperature field model output is the temperature distribution of each collection point in the greenhouse and its surrounding areas: Among them, T 全局 (P, t) is the predicted value of the global temperature field at spatial position P (x, y, z) and time t, is the Laplace operator of the temperature field, which indicates the degree of temperature diffusion in space. λ1, λ2, λ3 and λ4 are weights. G(x, y, z) is the spatial function of greenhouse environmental factors. H(t) is the function of environmental factors that change over time. is the time derivative of environmental factors, which is used to describe the immediate impact of environmental changes on the temperature field. is the time rate of change of the corrected temperature field, α2 is the coefficient controlling the temperature diffusion, To perform Fourier transform on the corrected temperature data.

4. According to claim 1, a method for local prediction of greenhouse temperature field based on differential steganographic modeling is characterized in that: The S4 comprises the following steps: S41. Determine the various areas inside the greenhouse k , the greenhouse is divided into several areas with different temperature requirements, and the temperature collection point position in each area is P k (x k ,y k ,z k ), where k represents the kth region, P k Indicates the temperature collection points in the area; S42. Collect temperature data in each area Perform differential processing to calculate the gradient change of the regional temperature field in, Represents the temperature field in region k at the spatial coordinate P k (x k ,y k ,z k ) to capture the temperature variation characteristics in the region with space; S43. According to the gradient change A regional temperature diffusion model is established, which is expressed as: Among them, α k is the temperature diffusion coefficient of region k, is the second-order derivative of the temperature field in the region, β k is the external factor influence coefficient of region k, F k (P k ,t) is the function of the external environmental factors of the region; S44. Combined with the temperature distribution characteristics within the region, the regional temperature field is expanded through steganographic modeling technology to generate a regional temperature field model T 区域 (P k ,t): Among them, λ k2 and λ k1 is the weight coefficient in the regional temperature field model, G k (x k ,y k ,z k ) and H k (t) represent the spatial and temporal environmental impact functions of region k respectively.

5. According to claim 1, a method for local prediction of greenhouse temperature field based on differential steganographic modeling is characterized in that: The S5 comprises the following steps: S51. Determine the key position inside the greenhouse and set the temperature collection point position at the micro scale as P m 9x m ,y m ,z m ), where m represents the identifier of the key position, P m Indicates the temperature collection point at micro scale; S52. Temperature data for key locations Perform differential processing to calculate the temperature field gradient at the microscopic scale in, It represents the gradient change of the temperature field at key positions at the microscopic scale, and is used to capture the temperature distribution of local details; S53. Based on gradient changes The temperature diffusion model at the microscopic scale is established and expressed as: Among them, α m is the temperature diffusion coefficient at the microscopic scale, is the second-order derivative of the temperature field, β m is the external factor influence coefficient, F m (P m ,t) is the function of external environmental factors at key locations; S54. Combine the temperature variation characteristics of key positions and use steganographic modeling technology to model the temperature field at the microscopic scale to generate a microscopic temperature field model T 微观 (P m ,t): Among them, λ m1 and λ m2 is the weight coefficient of the micro-scale model, G m (x m ,y m ,z m ) and H m (t) represent the spatial and temporal environmental impact functions of key locations respectively.

6. According to a method for local prediction of greenhouse temperature field based on differential steganographic modeling according to claim 1, it is characterized in that: The S6 comprises the following steps: S62. Fusion processing is performed on the global temperature field model, the regional temperature field model and the microscopic temperature field model to construct a comprehensive temperature field prediction model and set the weight parameters ω of the models at different scales. 全局 ,ω 区域 ,ω 微观 , generate the fused temperature field prediction data T 融合 (P,t): Among them, ω 全局 ,ω 区域 ,ω 微观 is the weight of models of different scales, indicating the contribution of each model to the overall temperature field prediction; is the second-order spatial derivative of the global temperature field to capture the global temperature diffusion; α 全局 is the parameter that controls the decay of the global temperature field over time; is the second-order derivative of the regional temperature field, combined with the regional environmental factor function G k (x k ,y k ,z k ) and time function H k The time derivative of (t); λ m is the adjustment coefficient of the micro-scale temperature field, is the second-order time derivative of the microscopic temperature field, δ m Control the time-varying decay of the microscopic temperature field, Ω, Ω k ,Ω m denote the spatial extents at the global, regional, and microscopic scales, respectively; S63. Comprehensive temperature field prediction data T 融合 (P, t) performs error analysis and correction by calculating the error ∈ 全局 ,∈ 区域 ,∈ 微观 The corrected temperature field prediction data is: Among them, ∈ 全局 ,∈ 区域 ,∈ 微观 are the error terms of the global, regional, and microscale models, respectively; S64. The corrected temperature field prediction data T 修正 (P, t) is stored as the final prediction data of the global and local temperature fields inside the greenhouse. The generated prediction data is represented by: T 最终 =[T 修正 (P1,t1),T 修正 (P2,t2),,T 修正 (P N ,t k )]; Among them, T 最终 Represents the matrix form of the final prediction data, P i is the location of the collection point inside the greenhouse, t k is each time point, and N is the number of sensors.

7. According to claim 1, a method for local prediction of greenhouse temperature field based on differential steganographic modeling is characterized in that: The S7 comprises the following steps: S71. Based on the corrected global and local temperature field prediction data T 修正 (P, t) Determine the temperature control requirements of each area inside the greenhouse, build the adjustment strategy of the temperature control system, and set the target temperature of the temperature control system as T 目标 (P,t); S72. Calculate the current actual temperature T in each area 当前 (P,t) and target temperature T 目标 The difference between (P, t) generates the temperature control adjustment instruction: ΔT 调节 (P,t)=T 目标 (P,t)-T 当前 (P,t); Where, ΔT 调节 (P,t) is the temperature adjustment required for each collection point, a positive value indicates that the temperature needs to be increased, and a negative value indicates that the temperature needs to be decreased; S73. Based on the temperature difference ΔT in different areas 调节 (P, t) By adjusting the heating or cooling equipment of the greenhouse temperature control system, the temperature of each area can be adjusted in real time; S74. Real-time monitoring of the adjustment effect of the temperature control system to obtain the adjusted temperature T 调节后 (P,t), and the target temperature T 目标 (P, t) for comparison, if the error ∈ 调节 If (P,t) exceeds the preset range, the temperature control strategy is automatically adjusted to correct the adjustment rate of the temperature control device; S75. Store the adjustment data of the temperature control system into the database, and record the temperature changes in each area, equipment adjustment parameters and adjustment effects.

Images