A precise intelligent agricultural greenhouse control system based on Zigbee transmission technology

Abstract

The application discloses a precise intelligent agricultural greenhouse control system based on Zigbee transmission technology and relates to the technical field of intelligent agricultural greenhouse environment control. The system continuously monitors leaf surface temperature, air temperature and humidity and breeze speed, identifies local dewing risks by combining the change track of the leaf surface temperature to the dew point temperature and the influence of the breeze speed on the leaf surface air boundary layer, constructs link congestion states by combining the node retransmission times and the cache queue occupation amount, and realizes the priority scheduling of the control instruction and the environment data transmission. In the case that the dewing risk is increased and the link is congested, the anti-dewing control instruction is preferentially issued to the local microcirculation fan and the top ventilation window, and the non-key sensing node report is dynamically suspended, so that the high precision, real-time performance and spatial positioning of the anti-dewing control are realized. The application can prevent the local low wind speed area from dewing in advance, improves the yield and quality of the greenhouse crops, and effectively reduces the local disease risk.

Description

Technical Field

[0001] This invention relates to the field of smart agricultural greenhouse environmental control technology, specifically to a precision smart agricultural greenhouse control system based on Zigbee transmission technology. Background Technology

[0002] Existing smart agricultural greenhouses typically connect air temperature and humidity sensors, leaf surface temperature sensors, wind speed sensors, and actuators to the greenhouse's main control unit via a Zigbee network, adjusting the greenhouse environment according to preset temperature and humidity thresholds. However, in scenarios involving continuous rain in winter and spring, nighttime irrigation, or prolonged periods of pre-sunrise insulation, low-wind-speed environments can easily form in localized areas near the north wall of the greenhouse, the ends of the greenhouse structure, and within the canopy of high-density crops. This causes leaf surface temperatures to rapidly approach dew point temperatures, resulting in localized condensation. Such condensation usually only occurs in a small number of planting areas and is difficult to detect promptly using average temperature and humidity within the greenhouse. Furthermore, in these localized environmental conditions... When the environment changes rapidly, multiple sensing nodes may report data to the Zigbee coordinator in a concentrated manner, leading to an increase in the number of data frame retransmissions, an increase in buffer queue usage, and the continuous occupation of transmission resources by non-critical periodic data. This, in turn, affects the timely transmission of critical environmental data and anti-condensation control information. In addition, existing control methods usually do not determine the priority control period based on the link congestion status, nor do they continuously assess the risk of local condensation based on updated environmental data during the control process and resume the periodic reporting of non-critical sensing nodes after the risk has decreased. Therefore, it is easy to cause problems such as delayed identification of local condensation, unreasonable allocation of communication resources, and insufficient real-time performance of anti-condensation control. Summary of the Invention

[0003] The purpose of this invention is to provide a precision intelligent agricultural greenhouse control system based on Zigbee transmission technology to address the shortcomings of the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a precision intelligent agricultural greenhouse control system based on Zigbee transmission technology, comprising: The inspection instruction parsing module obtains the anti-condensation inspection instructions issued by the greenhouse main control machine to determine the target greenhouse area, the target sensing node set, and the target execution node set. The environmental data acquisition module receives air temperature, relative humidity, leaf surface temperature, and micro-wind speed collected by the target sensing node set through the Zigbee coordinator, and records the number of data frame retransmissions and the amount of buffer queue occupied. The calculation module calculates the dew point temperature based on the air temperature and relative humidity, calculates the local condensation risk value based on the dew point temperature, leaf surface temperature and micro-wind speed, and calculates the link congestion value based on the number of data frame retransmissions and the buffer queue occupancy. The priority control window determination module determines the degree of condensation accumulation based on the difference between the local condensation risk value and the condensation exit threshold when the local condensation risk value reaches the condensation trigger threshold and the link congestion value reaches the congestion threshold. It then determines the priority control window based on the degree of condensation accumulation and the link congestion value. The condensation risk dynamic monitoring module collects updated environmental data of the target shed area in real time through the target sensing node set within the priority control window, and continuously updates the local condensation risk value based on the real-time collected updated environmental data. The anti-condensation control exit module exits anti-condensation control and resumes periodic reporting from non-critical sensing nodes when the updated local condensation risk value is lower than the condensation exit threshold.

[0005] Preferably, the inspection instruction parsing module includes: Receive the anti-condensation inspection command, and parse the shed area identifier, key sensing node list and execution node list from the anti-condensation inspection command; The target greenhouse area is determined by querying the greenhouse area mapping table based on the greenhouse area identifier; Based on the list of key sensing nodes and the list of execution nodes, the node management table and the execution device management table are queried respectively to determine the nodes that are associated with the target shed area and are online as the target sensing node set and the target execution node set.

[0006] Preferably, the environmental data acquisition module includes: A data acquisition request frame is generated based on the node address of each target sensing node in the target sensing node set, and the data acquisition request frame is sent to each target sensing node in a preset polling order. Receive environmental data frames returned by each target sensing node, perform node identity verification and timestamp marking on the environmental data frames, and accumulate the number of data frame retransmissions when any environmental data frame is not successfully received within a preset response time. The verified environment data frames are written into the cache queue, and the cache queue occupancy is determined based on the current number of data frames in the cache queue and the maximum number of data frames that the cache queue can hold.

[0007] Preferably, the computing module includes: Extract the air temperature and relative humidity corresponding to the target shed area from the environmental data frame; Based on the air temperature, the air relative humidity, the temperature correction factor, and the temperature correction constant, the temperature and humidity correlation parameters are determined according to the Magnus dew point calculation relationship. The dew point temperature is determined based on the temperature and humidity correlation parameters, the temperature correction coefficient, and the temperature correction constant.

[0008] Preferably, the calculation module further includes: Obtain the dew point temperature and leaf surface temperature at the current sampling time, and determine the current dew condensation approach state based on the temperature difference between the leaf surface temperature and the dew point temperature; The current dew-forming approach state is associated with the historical dew-forming approach states corresponding to multiple consecutive sampling periods to form a dew-forming approach trajectory that characterizes the continuous approach of the leaf surface temperature to the dew point temperature, and temperature difference convergence features are extracted based on the dew-forming approach trajectory. The air disturbance suppression level is determined based on the temperature difference convergence characteristics and the micro-wind speed changes within the corresponding sampling period, and the local condensation risk value is determined based on the temperature difference convergence characteristics and the air disturbance suppression level.

[0009] Preferably, the calculation module further includes: The number of data frame retransmissions, buffer usage time, and node identifiers of each target sensing node within the current acquisition window are associated to form a mapping of transmission conflict states between nodes. The inter-node transmission conflict state mapping is associated with the node reporting priority historical data to identify target sensing nodes with a high number of data frame retransmissions and a long buffer occupation time, and to determine the potential interference degree of each target sensing node to the transmission of key data frames. The link congestion value is determined based on the potential interference level, the transmission urgency of each target sensing node, and the fluctuation trend of the buffer queue occupancy.

[0010] Preferably, the priority control window determination module includes: When the local condensation risk value reaches the condensation trigger threshold and the link congestion value reaches the congestion threshold, the degree of condensation accumulation is determined based on the difference between the local condensation risk value and the condensation exit threshold, and the corresponding condensation reduction requirement level is determined based on the degree of condensation accumulation. The condensation reduction demand level is cross-matched with the link congestion value to determine the degree of resource bias in control command transmission relative to environmentally aware data transmission. Based on the resource allocation, the time ratios occupied by control command transmission and environmental perception data transmission within the subsequent acquisition window are reallocated, and the priority control window is determined based on the reallocated time ratios.

[0011] Preferably, the target execution node set includes at least a local micro-circulation fan and a top ventilation window. From the sensing nodes associated with the target shed area, sensing nodes not included in the target sensing node set are identified as non-critical sensing nodes, and a pause reporting instruction is sent to the non-critical sensing nodes to cause them to pause periodic reporting within the priority control window. Within the priority control window, the updated environmental data collected by the target sensing node set is received first, and a first-level control command is sent to the local micro-circulation fan. After the local micro-circulation fan is running, if the updated local condensation risk value is still not lower than the condensation exit threshold, a secondary control command is sent to the top ventilation window to open the top ventilation window at an opening degree corresponding to the degree of condensation accumulation.

[0012] Preferably, the dynamic monitoring module for condensation risk includes: Based on the actual installation location of each target sensing node in the target sensing node set, the target shed area is divided into multiple micro-environment grids, and each micro-environment grid corresponds to at least one target sensing node; Within the priority control window, the leaf surface temperature, dew point temperature, and micro-wind speed corresponding to each microenvironment grid are acquired respectively. The temperature difference approximation rate of each microenvironment grid is determined based on the temperature difference change between the leaf surface temperature and dew point temperature at adjacent sampling times, and the air retention degree of each microenvironment grid is determined based on the corresponding micro-wind speed. The corresponding grid risk weight is determined based on the temperature difference approach rate and air retention degree of each microenvironment grid, and the local condensation risk value is continuously updated based on the grid risk weight.

[0013] Preferably, the anti-condensation control exit module includes: When the updated local condensation risk value is lower than the condensation exit threshold, the local condensation risk value, leaf temperature and dew point temperature corresponding to multiple sampling times within the verification period are continuously acquired. When the proportion of valid sampling times lower than the condensation exit threshold to the total number of sampling times reaches more than 80%, and the leaf temperature at each sampling time is higher than the corresponding dew point temperature, the target greenhouse area is determined to enter a continuous non-condensation state. After the target shed area enters a continuous non-condensation state, an anti-condensation exit command is sent to the target execution node set, and the periodic reporting of non-critical sensing nodes is restored in batches according to the recorded order when the non-critical sensing nodes paused reporting. The number of nodes restored in each batch does not exceed 20% of the total number of nodes that paused reporting, and there is a collection window between two adjacent batches of nodes. After each batch of non-critical sensing nodes completes its periodic reporting and recovery, the link congestion value is re-determined based on the number of data frame retransmissions and the amount of buffer queue occupancy within the corresponding acquisition window. When the re-determined link congestion value is lower than the congestion threshold, the next batch of non-critical sensing nodes is recovered. When the re-determined link congestion value reaches the congestion threshold, the recovery of the next batch of non-critical sensing nodes is paused.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: The core technical approach of this invention lies in combining the continuous variation trajectory of leaf surface temperature and dew point temperature with the influence of micro-wind speed on the leaf surface air boundary layer to identify local condensation formation trends. This is then combined with the number of Zigbee node retransmissions and the amount of buffer queue usage to construct a link congestion state mapping. When the risk of local condensation increases and link congestion is significant, the time ratio of control commands to environmental data transmission is dynamically allocated to form a priority control window. Simultaneously, periodic reporting by non-critical nodes is suspended, prioritizing the transmission of critical control information. Unlike existing technologies that rely solely on average temperature and humidity or trigger fan opening based on single sampling, this invention, through the coupled analysis of continuous temperature difference changes and communication status, can identify condensation risks in local low-wind-speed areas in advance. It also ensures timely issuance of control commands even with limited critical communication resources, thereby accurately achieving local environmental regulation. The direct effect of this technical approach is that leaf condensation in local low-wind-speed areas of greenhouses can be predicted before reaching traditional humidity thresholds, significantly shortening the anti-condensation control response time and avoiding communication delays or control lags caused by concentrated node reporting. By restoring the periodic reporting of non-critical sensing nodes in batches, the smooth recovery of link resources is ensured, and the recurrence of instantaneous congestion is avoided. Furthermore, the dynamic micro-environment grid enables local condensation risk values ​​to accurately reflect spatial differences, ensuring precise positioning of fan and ventilation window controls. This achieves high precision and real-time performance in anti-condensation control, significantly reducing the risk of localized diseases and improving the yield and quality of greenhouse crops. The technical effect is significantly superior to existing control schemes based on average values ​​or single-point sampling. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 This is a flowchart of a precision smart agricultural greenhouse control system module based on Zigbee transmission technology according to the present invention.

[0017] Figure 2This is a flowchart of the dew point temperature generation method of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1, please refer to Figure 1 and Figure 2 As shown in this embodiment, a precision smart agricultural greenhouse control system based on Zigbee transmission technology includes: The inspection command parsing module obtains the anti-condensation inspection command issued by the greenhouse main control machine to determine the target greenhouse area, the target sensing node set, and the target execution node set.

[0020] On the greenhouse main control unit, anti-condensation inspection instructions are pre-generated. These instructions include the greenhouse area identification to be inspected, the corresponding list of key sensing nodes, and the list of execution nodes. They can be encapsulated in JSON or XML format, for example: { "shelf_id": "S01", "target_sensors": ["temp_01","humidity_01","leaf_01","wind_01"], "target_actuators": ["fan_01","vent_01"] } The Zigbee coordinator establishes a TCP / IP or serial communication connection with the host computer and receives inspection commands from the host computer periodically or on an event-triggered basis. Upon receiving a command, it parses the command content into a data object and verifies the integrity and validity of the fields, such as verifying the existence of the shed area identifier and the validity of the sensing node and execution node IDs.

[0021] The shelf_id field in the instruction corresponds to the greenhouse entity in the greenhouse database. The corresponding physical greenhouse information, including greenhouse layout, area and crop type, is matched by looking up a table.

[0022] Based on the target_sensors field, the coordinator selects air temperature and humidity sensors, leaf surface temperature sensors, and micro-wind speed sensors associated with the greenhouse area from the node management table to form a set of target sensing nodes for subsequent data collection.

[0023] Based on the target_actuators field, key execution nodes such as local micro-circulation fans and ventilation windows are selected from the execution equipment management table to form a target execution node set for subsequent control operations.

[0024] After the node set is completed, the coordinator performs status checks on each node, such as online status and the time of the last report, and initializes the node status to the pending inspection state to ensure that data can be collected and control commands can be issued in sequence within the inspection window.

[0025] The environmental data acquisition module receives air temperature, relative humidity, leaf surface temperature, and micro-wind speed collected by the target sensing node set through the Zigbee coordinator, and records the number of data frame retransmissions and the amount of buffer queue usage.

[0026] As one possible implementation, the environmental data acquisition module includes a Zigbee coordinator, a node management unit, and a link monitoring unit; wherein, air temperature and humidity sensing nodes, leaf surface temperature sensing nodes, and micro-wind speed sensing nodes are pre-deployed in the greenhouse, and each sensing node has a unique node address and is mapped to the corresponding greenhouse area.

[0027] After the inspection instruction parsing module determines the target sensing node set, the Zigbee coordinator generates a data acquisition request frame based on the node addresses in the target sensing node set, and sends acquisition instructions to each target sensing node in a preset polling order; after receiving the acquisition instructions, each target sensing node samples the current environmental parameters and encapsulates the air temperature, relative humidity, leaf surface temperature and micro-wind speed into an environmental data frame and returns it to the Zigbee coordinator.

[0028] After receiving the environmental data frame, the Zigbee coordinator timestamps the environmental data frame and verifies the node identity, and stores the verified environmental data in the cache queue. During the data reception process, the link monitoring unit monitors the transmission status of each environmental data frame in real time. When the same environmental data frame is not successfully received within the preset response time, a retransmission request is sent to the corresponding target sensing node, and the number of retransmissions is accumulated and recorded as the data frame retransmission count.

[0029] Meanwhile, the link monitoring unit monitors the data occupancy in the buffer queue in real time and calculates the buffer queue occupancy using the following formula: Q = Nq / Nmax; where Q is the buffer queue occupancy, Nq is the number of data frames in the current buffer queue, and Nmax is the maximum number of data frames that the buffer queue can hold.

[0030] Once all environmental data frames in the target sensing node set have been received, the environmental data acquisition module outputs air temperature, relative humidity, leaf surface temperature, wind speed, number of data frame retransmissions, and buffer queue occupancy, and sends them to the computing module for further processing.

[0031] The calculation module calculates the dew point temperature based on the air temperature and relative humidity, calculates the local condensation risk value based on the dew point temperature, leaf surface temperature and micro-wind speed, and calculates the link congestion value based on the number of data frame retransmissions and the buffer queue occupancy. As a preferred implementation method, the Magnus dew point calculation model, commonly used in the field of agricultural environmental monitoring, is used to calculate the dew point temperature. It has high calculation accuracy and is suitable for the temperature and humidity range of greenhouse environments.

[0032] After obtaining the air temperature and relative humidity, the calculation module first constructs the temperature and humidity correlation parameter α, and the calculation expression is: Where α is the temperature and humidity correlation parameter; RH is the relative humidity of the air, in %; T is the air temperature, in ℃; a is the temperature correction coefficient, taken as 17.27; b is the temperature correction constant, taken as 237.7; ln() is the natural logarithm function.

[0033] After obtaining the temperature and humidity correlation parameter α, the dew point temperature Td is calculated using the following expression: ;in, α is the dew point temperature in °C; α is the temperature-humidity correlation parameter; a is the temperature correction coefficient; b is the temperature correction constant.

[0034] Furthermore, when the air temperature is 25°C and the relative humidity is 80%: =1.4208; therefore, the dew point temperature is: ; In the current environment, when the leaf surface temperature drops to around 21.31℃, the leaf surface will reach the critical state of condensation.

[0035] Therefore, this embodiment calculates the dew point temperature using air temperature and relative humidity, providing a basic parameter for subsequent calculation of the risk of localized condensation. The dew point temperature reflects the temperature at which the air reaches saturation and can be used to characterize the critical conditions for condensation on leaves. This dew point temperature will serve as an input parameter for subsequent calculations of the difference between leaf temperature and dew point temperature to determine whether there is a risk of localized condensation in the target greenhouse area.

[0036] In this embodiment, the risk value of local condensation is not determined directly based on the leaf surface temperature and dew point temperature at a certain moment. Instead, it is calculated by combining the process of the leaf surface continuously approaching the dew point temperature and the influence of microwind speed on the renewal capacity of the leaf surface boundary layer, so as to improve the ability to identify the early stage of local condensation formation.

[0037] First, obtain the dew point temperature at the current sampling time. and leaf surface temperature And calculate the temperature difference between the two: ;in: This indicates the temperature difference at the current sampling time; This indicates the leaf surface temperature at the current sampling time, in °C. This represents the dew point temperature at the current sampling time, in °C. As the temperature gradually decreases, it indicates that the leaf surface temperature is approaching the dew point temperature, and the likelihood of condensation is gradually increasing. Subsequently, temperature difference data corresponding to multiple sampling periods are continuously acquired within a preset time window. Preferably, the sampling period is 30 seconds, and the time window is 10 minutes, thus obtaining a total of 20 sets of temperature difference data. To reflect the trend of the leaf surface temperature continuously approaching the dew point, the change in temperature difference between adjacent sampling periods was calculated: ;in, This represents the temperature difference change between the i-th sampling period and the (i+1)-th sampling period. Further, the average value of all temperature difference changes within the time window is calculated. The average value is used as a parameter to characterize the dew-forming trajectory.

[0038] When the average value is consistently greater than 0, it indicates that the leaf surface temperature is continuously approaching the dew point temperature; the larger the average value, the faster the approach.

[0039] To avoid misjudgment caused by fluctuations at individual sampling points, the dispersion of temperature difference changes within the time window was further calculated: ;in: This indicates the stability of temperature difference changes.

[0040] The smaller the SD, the more continuous and stable the process of the leaf surface approaching the dew point temperature, and the easier it is to form actual condensation.

[0041] At the same time, acquire micro-wind speed data within the corresponding sampling period: ; Calculate the average micro-wind speed within the time window .

[0042] Agricultural environmental monitoring studies have shown that when the wind speed inside the canopy is below 0.3 m / s, a stable moist boundary layer easily forms near the leaf surface, significantly reducing the efficiency of heat and water vapor exchange between the air and the leaf surface. Therefore, 0.3 m / s was selected as the reference wind speed in this embodiment.

[0043] Based on the reference wind speed, calculate the degree of air disturbance suppression: When the calculation result is less than 0, it is taken as 0. Wherein, K represents the degree of air disturbance suppression; the closer K is to 1, the weaker the air flow near the leaf surface; the closer K is to 0, the more sufficient the air exchange.

[0044] Subsequently, the local condensation risk value was calculated based on the condensation approach trajectory characterization parameters, the stability of temperature difference changes, and the degree of air disturbance suppression: Where R represents the local condensation risk value, Indicates the current temperature difference. This indicates the stability of temperature difference changes, K indicates the degree of suppression of air disturbances, and 0.1 is a correction factor used to avoid the denominator being 0.

[0045] From the above calculation process, we can see that: As the leaf surface temperature continues to approach the dew point temperature, the average temperature difference changes more, and the risk value increases. When the process of the leaf surface approaching the dew point temperature is relatively smooth, the stability of the temperature difference change decreases, and the risk value increases. When the wind speed decreases, making it difficult for the leaf boundary layer to be renewed, the degree of air disturbance suppression increases, and the risk value further increases. When the current temperature difference is very close to 0, the risk value increases rapidly.

[0046] For example, in the area along the north wall of the greenhouse before sunrise, continuous monitoring for 10 minutes showed that the leaf surface temperature dropped from 18.2℃ to 16.8℃, while the dew point temperature remained around 16.5℃. The average temperature difference change was 0.08℃ / cycle, the temperature difference stability was 0.02℃, and the average gale speed was 0.05m / s. Therefore, the air disturbance suppression level was 0.83, and the current temperature difference was 0.3℃. Substituting these values ​​into the above calculation process yielded a high risk value for localized condensation. Although the leaf surface temperature had not yet reached the dew point temperature, the system could already identify that the leaves were continuously approaching the condensation state, thus initiating the subsequent control process ahead of schedule.

[0047] Compared to a single-point comparison using only leaf surface temperature and dew point temperature, this embodiment uses the dew condensation trajectory to reflect the continuous evolution of leaf surface temperature, and combines the influence of micro-wind speed on the leaf boundary layer renewal capacity for correction. This can identify local dew condensation trends in advance, reduce misjudgments caused by fluctuations in a single sampling, and improve the accuracy and lead time of local dew condensation risk assessment.

[0048] In this embodiment, the calculation of link congestion value does not simply rely on the average number of retransmissions or instantaneous buffer usage. Instead, it is comprehensively quantified by establishing a mapping of transmission conflict states between nodes and combining it with historical data on node reporting priorities, so as to accurately reflect the pressure state of the Zigbee communication channel within the current collection window.

[0049] First, within the current acquisition window, record the number of data frame retransmissions for each node in the target sensing node set. and the buffer duration occupied by each data frame in the coordinator buffer queue Then, associate each node with its corresponding node identifier. Construct a mapping table of transmission conflict states between nodes, where each record includes the node identifier, number of retransmissions, buffer usage time, and the time of the most recent report.

[0050] Subsequently, the transmission conflict state mapping table was correlated with historical node reporting priority data for analysis. This historical priority data includes the node's average reporting latency and the frequency of participation in critical control events within past acquisition windows, used to determine the potential interference level of nodes on critical data transmission. A potential interference index was calculated by weighting the number of retransmissions and buffer usage time for each node. The expression is: ;in, This represents the potential interference index of the i-th node. This represents the maximum possible number of retransmissions for a node, used for normalization. This is the maximum allowed time for the cache queue to be occupied.

[0051] By combining the potential interference indices of all nodes with the urgency of node transmission, a comprehensive link congestion value R is generated: Where n is the number of target sensing nodes; R is the transmission urgency index for the i-th node in the current acquisition window, and is assigned a value based on the frequency of the node's participation in critical control events (1 for high-frequency nodes and 0.5 for low-frequency nodes). The larger R is, the higher the current Zigbee link congestion level, and the more likely the transmission of critical data frames may be blocked.

[0052] For example, if a target shed area contains 10 nodes, and 3 of these nodes retransmit more than 5 times consecutively with buffer usage exceeding 0.8 seconds, and their urgency level is 1, then the calculated link congestion value is significantly higher than that of the other nodes, thus reflecting the communication channel pressure in advance. This method allows for the prediction of potential congestion when high-density nodes report data, enabling the pre-allocation of control command transmission windows in the next stage of condensation prevention and control, ensuring the real-time transmission of critical control data and the effectiveness of localized condensation risk mitigation.

[0053] The priority control window determination module determines the degree of condensation accumulation based on the difference between the local condensation risk value and the condensation exit threshold when the local condensation risk value reaches the condensation trigger threshold and the link congestion value reaches the congestion threshold. It then determines the priority control window based on the degree of condensation accumulation and the link congestion value.

[0054] In this embodiment, to ensure timely response of local condensation control under Zigbee link congestion, the degree of condensation accumulation is calculated by the difference between the local condensation risk value and the condensation exit threshold, and control resources are dynamically allocated in combination with the link congestion value to determine the priority control window.

[0055] It should be noted that the dew point exit threshold is used to characterize a state where the risk of localized dew point condensation has decreased to the point where continued anti-dew point control is no longer necessary. Its selection is based on the safe temperature difference between leaf surface temperature and dew point temperature, as well as the actual changing patterns of the local microclimate environment within the greenhouse. Continuous monitoring of different crop planting areas revealed that when the leaf surface temperature is higher than the dew point temperature and maintains a certain safety margin, the leaf surface no longer possesses the thermodynamic conditions for continuous condensation of water vapor, and localized dew point condensation gradually disappears. Based on this, statistical analysis was performed on the localized dew point condensation risk values ​​corresponding to the disappearance of dew point condensation from a large amount of measured data. The risk value covering more than 95% of the samples in the non-dew point state was selected as the dew point exit threshold, ensuring that anti-dew point control is only discontinued after the leaf surface has moved out of the dew point danger zone. This setting method avoids premature exit leading to re-condensation in localized areas and also prevents energy waste caused by prolonged continuous operation of fans and ventilation equipment, establishing a stable correlation between the dew point exit threshold, the effectiveness of anti-dew point control, and equipment operating efficiency.

[0056] The congestion threshold is used to determine whether there is a transmission bottleneck in the Zigbee communication channel, thus affecting the real-time issuance of critical anti-condensation control commands. Its selection is based on actual observation data of node retransmission counts, buffer queue occupancy, and node reporting concentration. By statistically analyzing the distribution of nodes with high retransmission and long buffer occupancy within a historical acquisition window, the risk level of potential delays in critical command transmission under this condition is determined. The design principle is to ensure that the transmission of critical control data is prioritized when the link experiences high load or concentrated node reporting, rather than relying solely on the average frame loss rate. The determination method includes: firstly, collecting the number of data frame retransmissions and buffer occupancy time for each node within a continuous acquisition window, and weighting and accumulating these data based on the historical urgency of node reporting to form a comprehensive congestion index; then, adjusting this index in an experimental environment to ensure that more than 95% of critical control frames are transmitted on time even under high load scenarios, and setting the corresponding index value as the congestion threshold. This threshold enables real-time assessment of communication channel pressure. When the congestion value reaches or exceeds the threshold, the system automatically adjusts the priority of data acquisition and control commands, thereby significantly improving the real-time performance and effectiveness of local condensation control.

[0057] First, based on the local condensation risk value R and the condensation exit threshold within the current collection window. Calculate the degree of condensation accumulation C: Where C represents the current level of condensation accumulation, used to quantify the degree to which localized condensation on the leaf surface deviates from the exit conditions; This indicates the threshold for condensation to exit.

[0058] Based on the degree of condensation accumulation (C), the local condensation control requirements are divided into three levels: High grade: C≥0.6, indicating that the leaf surface temperature is continuously approaching the dew point and the risk of condensation is rapidly increasing; Medium range: 0.3≤C<0.6, indicating that the leaf surface temperature is clearly approaching the dew point; Low level: 0≤C<0.3, indicating that the leaf surface temperature is close to but has not yet entered a high-risk state.

[0059] Subsequently, the condensation reduction demand level is cross-matched with the link congestion value L. The link congestion value L is calculated using the aforementioned node retransmission count and buffer usage, reflecting the current pressure status of the communication channel. The cross-matching uses a weighted summation method to obtain the resource tilt coefficient S: S = 0.7 × C + 0.3 × L; where S represents the degree of resource tilt, used for the allocation of subsequent control command and sensing data transmission time.

[0060] After obtaining the resource tilt coefficient S, the time allocated for environmental perception data transmission and control command transmission within the subsequent acquisition window is dynamically reallocated. Specifically, this is implemented as follows: Control command time ratio =S×100%, when S is 1, control command transmission occupies the entire priority window; the proportion of sensor data transmission occupies... .

[0061] Finally, the duration of the priority control window is determined based on the proportion of control command transmissions. The expression is: ;in, The total duration of the preset acquisition window is designed based on the greenhouse environment monitoring cycle and Zigbee frame interval, and is set to 60 seconds in this embodiment. The window duration is controlled first and dynamically adjusted according to the degree of condensation accumulation and the link congestion value.

[0062] In this embodiment, the above method ensures that when the leaf surface temperature is constantly approaching the dew point and the link congestion is high, the control command transmission prioritizes the use of communication resources, thereby reducing the risk of local condensation in a timely manner. At the same time, the sensing data transmission is retained to update the risk status, achieving high-precision and real-time anti-condensation control.

[0063] The condensation risk dynamic monitoring module collects updated environmental data of the target shed area in real time through the target sensing node set within the priority control window, and continuously updates the local condensation risk value based on the real-time collected updated environmental data.

[0064] As one implementation method, in order to avoid the local abnormal areas being averaged out by using a single sampling point to update the local condensation risk value within the priority control window, this embodiment uses the continuously collected updated environmental data from the target sensing node set to construct a dynamic microenvironment grid, and continuously corrects the local condensation risk value based on the dynamic microenvironment grid.

[0065] Specifically, the leaf surface temperature, dew point temperature, and micro-wind speed corresponding to each target sensing node within the target greenhouse area are obtained, and a spatial coordinate set is established according to the actual installation location of the nodes. Based on the distance between adjacent target sensing nodes, the target greenhouse area is divided into multiple micro-environment grids, each micro-environment grid corresponding to a local monitoring area. For the i-th micro-environment grid, its current grid temperature difference is calculated: ;in, Let be the current grid temperature difference of the i-th microenvironment grid; Let i be the leaf surface temperature corresponding to the i-th microenvironment grid. Let be the dew point temperature corresponding to the i-th microenvironment grid.

[0066] Subsequently, during continuous sampling, the historical grid temperature difference corresponding to each microenvironment grid is recorded. Let the current sampling time be t, then the temperature difference change between two adjacent sampling times is calculated: ;when When the value remains above 0, it indicates that the leaf surface temperature is continuously approaching the dew point temperature.

[0067] Furthermore, the average value of all temperature difference changes within the preset analysis period is statistically analyzed to obtain the temperature difference approximation rate: ;in, Let m be the temperature difference approximation rate of the i-th microenvironment grid, and m be the number of samples within the analysis period.

[0068] At the same time, the micro-wind speed sequence corresponding to the i-th micro-environment grid is obtained, and the average micro-wind speed is calculated: ;in, Let be the average micro-wind speed of the i-th microenvironment grid. The wind speed corresponds to the kth sampling.

[0069] Because of the boundary layer phenomenon near the leaves in agricultural greenhouses, when the wind speed decreases, the air renewal capacity around the leaves decreases, and the water vapor released from the leaves is not easily dispersed. Therefore, it is necessary to further calculate the degree of air retention. The value is 0 when the calculation result is less than 0. Wherein, The degree of air retention; The reference wind speed is 0.3 m / s, based on data from environmental control experiments in facility agriculture. Numerous experiments have shown that when the wind speed inside the crop canopy exceeds 0.3 m / s, the air exchange capacity near the leaf surface is significantly enhanced, and localized condensation is significantly reduced; when the wind speed is below 0.3 m / s, the leaf boundary layer tends to stabilize, and the probability of condensation increases rapidly. Therefore, 0.3 m / s is selected as the boundary reference value for air exchange.

[0070] After obtaining the temperature difference approximation rate and air retention rate, the risk weights of the microenvironment grid are calculated: ;in, 0.1 represents the risk weight of the microenvironment grid; 0.1 is a correction amount used to avoid outliers when the current grid temperature difference is close to 0.

[0071] Subsequently, the local condensation risk value is updated based on the risk weight of each microenvironment grid: ;in, The updated risk value for localized condensation. is the local condensation risk value corresponding to the previous sampling period; n is the number of microenvironment grids.

[0072] For example, four target sensing nodes were deployed near the north wall of the target greenhouse area, resulting in four microenvironment grids. In the third microenvironment grid, the leaf surface temperature continuously decreased, with the current grid temperature difference dropping from 0.8℃ to 0.2℃, and the average micro-wind speed was only 0.06 m / s, while the average micro-wind speeds of the other microenvironment grids were all greater than 0.25 m / s. At this time, the temperature difference approach rate and air retention degree corresponding to the third microenvironment grid were significantly higher than those of other areas, and its risk weight would increase significantly, thereby increasing the updated local condensation risk value.

[0073] In this embodiment, instead of using the average environmental parameters of the target greenhouse area to update the local condensation risk value, the dynamic micro-environment grid is used to identify the process of the leaf surface temperature continuously approaching the dew point temperature in the local low wind speed area. The degree of air retention reflects the renewal capacity of the leaf boundary layer, so that the updated local condensation risk value can more accurately reflect the true condensation state of the local area, thereby improving the pertinence and real-time performance of subsequent anti-condensation control.

[0074] The anti-condensation control exit module exits anti-condensation control and resumes periodic reporting from non-critical sensing nodes when the updated local condensation risk value is lower than the condensation exit threshold.

[0075] As one implementation method, after continuously updating the local condensation risk value within the priority control window, when the updated local condensation risk value is lower than the condensation exit threshold, the anti-condensation control is not immediately exited. Instead, the stability of the local condensation risk value is first verified to avoid false exit due to a temporary decrease in the local condensation risk value caused by instantaneous environmental fluctuations.

[0076] Specifically, the local condensation risk value corresponding to multiple sampling times within a preset verification period is continuously acquired, and the number of valid samplings that meet the condition of being below the condensation exit threshold is counted. When the proportion of valid samplings to the total number of samplings reaches more than 80%, it is determined that the target shed area has entered a continuous non-condensation state.

[0077] Subsequently, the leaf surface temperature and dew point temperature corresponding to the verification period are obtained, and the temperature difference between the leaf surface temperature and the dew point temperature is calculated; when the temperature difference corresponding to all sampling times is greater than 0, it is determined that the leaf surface has escaped the conditions for condensation formation, thereby generating an anti-condensation exit command.

[0078] After generating the anti-condensation exit command, non-critical sensing nodes are resumed reporting in batches according to the recorded order when the nodes paused reporting. The number of nodes resumed in each batch does not exceed 20% of the total number of paused nodes, and there is a data collection window interval between adjacent batches. After each batch of nodes is resumed, the number of data frame retransmissions and the buffer queue usage within the corresponding data collection window are recorded, and it is determined whether the link congestion value remains below the congestion threshold.

[0079] When the link congestion value remains below the congestion threshold, continue to restore the next batch of non-critical sensing nodes; when the link congestion value approaches the congestion threshold again, pause the subsequent node restoration operation, and continue to run the current number of restored nodes until the link congestion value drops again, then continue to restore the remaining nodes.

[0080] When all non-critical sensing nodes resume periodic reporting and the local condensation risk value within multiple consecutive collection windows is lower than the condensation exit threshold, the current condensation prevention control process ends, and the normal environmental monitoring status of the target shed area is restored.

[0081] Compared to the method of immediately restoring all node reporting after the condensation risk value decreases, this invention can avoid the instantaneous congestion of the Zigbee link caused by a large number of nodes resuming communication at the same time. At the same time, it uses the continuous separation state of leaf surface temperature and dew point temperature to verify whether local condensation has been truly eliminated, thereby improving the reliability of the anti-condensation control exit and ensuring that the communication link smoothly returns to normal operation.

[0082] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A precision smart agricultural greenhouse control system based on Zigbee transmission technology, characterized in that, include: The inspection instruction parsing module obtains the anti-condensation inspection instructions issued by the greenhouse main control machine to determine the target greenhouse area, the target sensing node set, and the target execution node set. The environmental data acquisition module receives air temperature, relative humidity, leaf surface temperature, and micro-wind speed collected by the target sensing node set through the Zigbee coordinator, and records the number of data frame retransmissions and the amount of buffer queue occupied. The calculation module calculates the dew point temperature based on the air temperature and relative humidity, calculates the local condensation risk value based on the dew point temperature, leaf surface temperature and micro-wind speed, and calculates the link congestion value based on the number of data frame retransmissions and the buffer queue occupancy. The priority control window determination module determines the degree of condensation accumulation based on the difference between the local condensation risk value and the condensation exit threshold when the local condensation risk value reaches the condensation trigger threshold and the link congestion value reaches the congestion threshold. It then determines the priority control window based on the degree of condensation accumulation and the link congestion value. The condensation risk dynamic monitoring module collects updated environmental data of the target shed area in real time through the target sensing node set within the priority control window, and continuously updates the local condensation risk value based on the real-time collected updated environmental data. The anti-condensation control exit module exits anti-condensation control and resumes periodic reporting from non-critical sensing nodes when the updated local condensation risk value is lower than the condensation exit threshold.

2. The precision intelligent agricultural greenhouse control system based on Zigbee transmission technology according to claim 1, characterized in that, The inspection instruction parsing module includes: Receive the anti-condensation inspection command, and parse the shed area identifier, key sensing node list and execution node list from the anti-condensation inspection command; The target greenhouse area is determined by querying the greenhouse area mapping table based on the greenhouse area identifier; Based on the list of key sensing nodes and the list of execution nodes, the node management table and the execution device management table are queried respectively to determine the nodes that are associated with the target shed area and are online as the target sensing node set and the target execution node set.

3. The precision intelligent agricultural greenhouse control system based on Zigbee transmission technology according to claim 2, characterized in that, The environmental data acquisition module includes: A data acquisition request frame is generated based on the node address of each target sensing node in the target sensing node set, and the data acquisition request frame is sent to each target sensing node in a preset polling order. Receive environmental data frames returned by each target sensing node, perform node identity verification and timestamp marking on the environmental data frames, and accumulate the number of data frame retransmissions when any environmental data frame is not successfully received within a preset response time. The verified environment data frames are written into the cache queue, and the cache queue occupancy is determined based on the current number of data frames in the cache queue and the maximum number of data frames that the cache queue can hold.

4. A precision smart agricultural greenhouse control system based on Zigbee transmission technology according to claim 3, characterized in that, The computing module includes: Extract the air temperature and relative humidity corresponding to the target shed area from the environmental data frame; Based on the air temperature, the air relative humidity, the temperature correction factor, and the temperature correction constant, the temperature and humidity correlation parameters are determined according to the Magnus dew point calculation relationship. The dew point temperature is determined based on the temperature and humidity correlation parameters, the temperature correction coefficient, and the temperature correction constant.

5. A precision smart agricultural greenhouse control system based on Zigbee transmission technology according to claim 4, characterized in that, The computing module also includes: Obtain the dew point temperature and leaf surface temperature at the current sampling time, and determine the current dew condensation approach state based on the temperature difference between the leaf surface temperature and the dew point temperature; The current dew-forming approach state is associated with the historical dew-forming approach states corresponding to multiple consecutive sampling periods to form a dew-forming approach trajectory that characterizes the continuous approach of the leaf surface temperature to the dew point temperature, and temperature difference convergence features are extracted based on the dew-forming approach trajectory. The air disturbance suppression level is determined based on the temperature difference convergence characteristics and the micro-wind speed changes within the corresponding sampling period, and the local condensation risk value is determined based on the temperature difference convergence characteristics and the air disturbance suppression level.

6. A precision smart agricultural greenhouse control system based on Zigbee transmission technology according to claim 5, characterized in that, The computing module also includes: The number of data frame retransmissions, buffer usage time, and node identifiers of each target sensing node within the current acquisition window are associated to form a mapping of transmission conflict states between nodes. The inter-node transmission conflict state mapping is associated with the node reporting priority historical data to identify target sensing nodes with a high number of data frame retransmissions and a long buffer occupation time, and to determine the potential interference degree of each target sensing node to the transmission of key data frames. The link congestion value is determined based on the potential interference level, the transmission urgency of each target sensing node, and the fluctuation trend of the buffer queue occupancy.

7. A precision smart agricultural greenhouse control system based on Zigbee transmission technology according to claim 6, characterized in that, The priority control window determination module includes: When the local condensation risk value reaches the condensation trigger threshold and the link congestion value reaches the congestion threshold, the degree of condensation accumulation is determined based on the difference between the local condensation risk value and the condensation exit threshold, and the corresponding condensation reduction requirement level is determined based on the degree of condensation accumulation. The condensation reduction demand level is cross-matched with the link congestion value to determine the degree of resource bias in control command transmission relative to environmentally aware data transmission. Based on the resource allocation, the time ratios occupied by control command transmission and environmental perception data transmission within the subsequent acquisition window are reallocated, and the priority control window is determined based on the reallocated time ratios.

8. A precision smart agricultural greenhouse control system based on Zigbee transmission technology according to claim 1, characterized in that, The target execution node set includes at least local micro-circulation fans and top ventilation windows. From the sensing nodes associated with the target shed area, sensing nodes not included in the target sensing node set are identified as non-critical sensing nodes, and a pause reporting instruction is sent to the non-critical sensing nodes to cause them to pause periodic reporting within the priority control window. Within the priority control window, the updated environmental data collected by the target sensing node set is received first, and a first-level control command is sent to the local micro-circulation fan. After the local micro-circulation fan is running, if the updated local condensation risk value is still not lower than the condensation exit threshold, a secondary control command is sent to the top ventilation window to open the top ventilation window at an opening degree corresponding to the degree of condensation accumulation.

9. A precision smart agricultural greenhouse control system based on Zigbee transmission technology according to claim 8, characterized in that, The dynamic monitoring module for condensation risk includes: Based on the actual installation location of each target sensing node in the target sensing node set, the target shed area is divided into multiple micro-environment grids, and each micro-environment grid corresponds to at least one target sensing node; Within the priority control window, the leaf surface temperature, dew point temperature, and micro-wind speed corresponding to each microenvironment grid are acquired respectively. The temperature difference approximation rate of each microenvironment grid is determined based on the temperature difference change between the leaf surface temperature and dew point temperature at adjacent sampling times, and the air retention degree of each microenvironment grid is determined based on the corresponding micro-wind speed. The corresponding grid risk weight is determined based on the temperature difference approach rate and air retention degree of each microenvironment grid, and the local condensation risk value is continuously updated based on the grid risk weight.

10. A precision smart agricultural greenhouse control system based on Zigbee transmission technology according to claim 9, characterized in that, The anti-condensation control exit module includes: When the updated local condensation risk value is lower than the condensation exit threshold, the local condensation risk value, leaf temperature and dew point temperature corresponding to multiple sampling times within the verification period are continuously acquired. When the proportion of valid sampling times lower than the condensation exit threshold to the total number of sampling times reaches more than 80%, and the leaf temperature at each sampling time is higher than the corresponding dew point temperature, the target greenhouse area is determined to enter a continuous non-condensation state. After the target shed area enters a continuous non-condensation state, an anti-condensation exit command is sent to the target execution node set, and the periodic reporting of non-critical sensing nodes is restored in batches according to the recorded order when the non-critical sensing nodes paused reporting. The number of nodes restored in each batch does not exceed 20% of the total number of nodes that paused reporting, and there is a collection window between two adjacent batches of nodes. After each batch of non-critical sensing nodes completes its periodic reporting and recovery, the link congestion value is re-determined based on the number of data frame retransmissions and the amount of buffer queue occupancy within the corresponding acquisition window. When the re-determined link congestion value is lower than the congestion threshold, the next batch of non-critical sensing nodes is recovered. When the re-determined link congestion value reaches the congestion threshold, the recovery of the next batch of non-critical sensing nodes is paused.

Images