A remote control and intelligent adjustment method based on intelligent irrigation electric valve

Abstract

The application discloses a kind of remote control and intelligent regulation method based on intelligent irrigation electric valve, it is related to intelligent irrigation control technical field, comprising the following steps: access irrigation environment state, pipe network hydraulic state and valve execution state, generate irrigation state sequence;Irrigation state sequence is input root layer hydrodynamic mixed prediction model, generate root layer predicted humidity profile and water demand rate curve;Based on water demand priority and pipe network hydraulic state generate pipe network capacity check result;Electric valve target opening is compensated and limited, and S-shaped speed curve is dynamically adjusted, and step motor pulse frequency and PWM drive instruction are generated;Valve control instruction is generated after cloud communication timeout;According to valve opening and closing constraint parameter, generate limited microstep regulation instruction;Based on irrigation execution feedback update model parameter and control rule.The application improves the stability, regulation accuracy and emergency control capability of irrigation valve remote control.

Description

Technical Field

[0001] This invention relates to the field of intelligent irrigation control technology, and in particular to a remote control and intelligent adjustment method based on an intelligent irrigation electric valve. Background Technology

[0002] With the development of agricultural water-saving irrigation and IoT remote control technologies, intelligent irrigation systems based on sensor acquisition, cloud platform analysis, and electric valve execution have been widely proposed. Among existing technologies, CN111280034A discloses irrigation action control through a sensor array, remote communication module, and valve drive system; CN110896831A discloses adjusting the opening degree of an electric regulating valve based on soil moisture, flow rate, and pressure feedback; CN117390968B discloses fusing Richards equations with a physical information neural network for soil water movement modeling; and CN113721686A discloses field terminals, a central server, and offline self-control capabilities in a smart agriculture monitoring system. These examples demonstrate that remote valve control, pressure and flow feedback, soil moisture physical modeling, and edge automation all have publicly available foundations.

[0003] However, existing intelligent irrigation systems still face significant challenges in large-scale farmland scenarios: First, soil water demand assessment often relies on single-point humidity, infiltration rate, or independent soil moisture models, making it difficult to directly reflect root zone water shortage and historical irrigation residuals. Second, valve target opening is typically determined based on single-valve demand or pressure-flow feedback, lacking a mechanism to uniformly verify root zone water demand priority, network capacity, and water hammer risk, which can easily lead to multiple valves competing for water, insufficient end-point pressure, and pipe burst risks. Third, valve execution often employs fixed opening, mechanical deceleration, or conventional feedback adjustment, making it difficult to dynamically adjust the opening and closing curves based on water hammer risk constraints and historical deviations. Fourth, automatic control during network outages often relies on local preset strategies, failing to reuse the most recent effective network capacity verification results, which can easily lead to blind valve opening, valve stopping, or irrigation interruption in weak network environments.

[0004] Therefore, how to provide a remote control and intelligent regulation method based on intelligent irrigation electric valves is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] One objective of this invention is to propose a remote control and intelligent adjustment method based on intelligent irrigation electric valves. This invention fully utilizes root layer hydrodynamic mixing prediction, pipeline capacity verification, valve opening and closing constraint control, and edge emergency control technologies. It details the closed-loop control process from irrigation status perception, water demand prediction, target opening generation, S-shaped velocity curve driving, restricted micro-step adjustment to execution feedback update. It has the advantages of high irrigation water demand matching, good valve adjustment stability, low pipeline water hammer risk, and strong emergency response capability in case of network failure.

[0006] A remote control and intelligent adjustment method based on an intelligent irrigation electric valve according to an embodiment of the present invention includes the following steps: Step 1: Access the irrigation environment status, pipeline hydraulic status, and valve execution status, and generate an irrigation status sequence through edge gateway processing; Step 2: Input the irrigation state sequence into the root zone hydrodynamic mixed prediction model, generate the water transport baseline trajectory based on soil hydraulic parameters through the Richards equation, generate residual memory features through the RWKV-TS network based on the consistency of historical irrigation state and physical residual, and perform constraint correction according to the root zone water shortage state by root zone stress gating, and generate the root zone predicted humidity profile and water demand rate curve. Step 3: Determine water demand priority based on the predicted humidity profile and water demand rate curve of the root layer, and generate network capacity verification results including target opening degree of electric valves and water hammer risk constraints in combination with the hydraulic state of the pipeline network; Step 4: Based on the pipeline capacity verification results, valve flow characteristics and valve execution status, compensate and limit the target opening of the electric valve, generate valve opening and closing constraint parameters, and dynamically adjust the S-shaped speed curve according to the water hammer risk constraint and valve opening and closing constraint parameters to generate stepper motor pulse frequency and PWM drive command. Step 5: After the cloud communication status meets the preset communication timeout conditions, read the most recent effective pipeline capacity verification result and irrigation status sequence, and generate valve control commands according to the edge emergency control rules; Step 6: Generate restricted micro-step adjustment commands based on valve opening and closing constraint parameters, valve execution status, and network hydraulic status; Step 7: Generate irrigation execution feedback based on valve control commands, restricted microstep adjustment commands, and post-irrigation state changes. Update the root layer hydrodynamic mixing prediction model parameters, pipeline capacity verification rules, valve opening and closing constraint parameters, and edge emergency control rules based on the irrigation execution feedback.

[0007] Optionally, step one specifically includes: Establish the binding relationship between irrigation control zones, pipeline branches, and electric valves based on the irrigation control zone configuration table; The system accesses the irrigation environment status via sensor interface, the hydraulic status of the pipeline network via pipeline network monitoring interface, and the valve execution status via valve controller communication interface. Perform field encoding matching, status value extraction, and dimension conversion on irrigation environment status, pipeline hydraulic status, and valve execution status to generate environment status field, hydraulic status field, and execution status field; The acquisition time is corrected using the edge gateway system clock as a reference, and a control cycle identifier is generated according to the preset control cycle. Perform invalid binding removal, value range verification, adjacent control cycle mutation verification, and missing field marking on the environmental status field, hydraulic status field, and execution status field; The fields that have completed status cleaning are matched according to the irrigation control area number, pipeline branch number and control cycle identifier to generate status combination records, and then combined according to the time order of the control cycle identifier to generate an irrigation status sequence.

[0008] Optionally, step two specifically includes: The root zone hydrodynamic mixing prediction model includes a Richards equation solution layer, an RWKV-TS network, and root zone stress gating. Historical irrigation states are formed based on root layer humidity, irrigation input, evapotranspiration and valve control results in the irrigation state sequence. Soil moisture characteristic curves and unsaturated hydraulic conductivity curves are established based on soil hydraulic parameters, and the root layer is divided into multiple water transport calculation layers. The Richards equation solution layer uses a finite volume discretization method, taking each water transport calculation layer as a water balance control volume, and adopts an implicit iteration method, taking the water content of each water transport calculation layer in the next control period as the solution quantity, and calculating the water balance residual based on infiltration flux, drainage flux, root water uptake, interlayer water flux and water content change. When the water balance residual exceeds the preset residual threshold, the infiltration flux or root water absorption term is corrected according to the direction of the water balance residual and the solution is iterated again until the water balance residual meets the preset residual threshold, and a water transport baseline trajectory is generated. Physical residuals are generated based on the difference between the water transport baseline trajectory and the root layer humidity state, and physical residual consistency is generated according to the continuous same-direction accumulation, sign reversal and abrupt change states of physical residuals. The RWKV-TS network processes historical irrigation states, water transport baseline trajectories, and physical residual consistency through a temporal mixing layer, a channel mixing layer, a recursive state buffer layer, and a residual output layer. The temporal mixing layer generates key vectors, value vectors, and gating vectors based on state embedding. The recursive state buffer layer generates exponentially decaying weights based on the control period interval and a learnable time constant, and adjusts the degree of historical residual retention according to physical residual consistency. The residual output layer generates residual memory features. The root zone stress gating generates a modified gating coefficient based on the water shortage state of the root layer. It performs gating processing on the water content increase direction correction component and water content decrease direction correction component in the residual memory characteristics, and performs field capacity limit processing on the single layer water demand correction, generating a predicted humidity profile and water demand rate curve for the root layer.

[0009] Optionally, step three specifically includes: The water-deficient layer is determined based on the water transport calculation layer below the target moisture state limit of the crop in the root layer predicted humidity profile, and the root zone water demand record is generated by combining the cumulative water deficit of the water-deficient layer and the water demand rate curve. Water demand priority is determined based on whether the water-deficient layer includes the preset crop taproot zone, the cumulative water deficit of the water-deficient layer, and the trend of water demand rate curve. A water demand queue is generated based on the cumulative water demand, the predicted humidity of the taproot zone, and the pressure margin of the water supply path. The water supply path is determined based on the binding relationship between irrigation control areas, pipeline branches and electric valves, and the water demand queuing sequence is allocated with capacity constraints based on the pressure margin of the water supply path and the available water supply capacity, generating the allocated flow or the suspended waiting state. The target opening degree of the electric valve is determined based on the allocated flow rate, the current pipeline pressure status, and the valve flow characteristics. The water hammer risk constraint is generated based on the change in the target opening degree, the expected change in flow rate, and the pressure margin of the water supply path. The pipeline capacity verification result is generated by combining the water demand queuing sequence, allocated flow rate, target opening degree of electric valves, hanging waiting status, and water hammer risk constraints.

[0010] Optionally, step four specifically includes: The valve flow sensitivity corresponding to the target opening degree of the electric valve is determined based on the valve flow characteristics, and the opening compensation amount is generated based on the continuous deviation between the target opening degree and the actual opening degree to compensate for the target opening degree of the electric valve. Based on the allocated flow rate, the allocated flow rate is limited to the compensated target opening. Based on the water hammer risk constraint and the water supply path pressure margin, the allowable flow rate change boundary is determined. The water hammer risk is limited to the compensated target opening to obtain the target opening. Valve opening and closing constraint parameters are generated based on the target opening degree, opening degree adjustment direction, single opening degree adjustment amount, valve flow sensitivity, allowable flow change boundary and water hammer risk constraint. Using water hammer risk constraints, single opening adjustment, valve flow sensitivity, and historical opening deviation as indexes, adjacent parameter points are selected in the preset S-shaped speed curve parameter calibration table and interpolated to generate the speed limit, acceleration segment duration, constant speed segment duration, deceleration segment duration, end buffer opening range, and pulse frequency change limit. Based on the above parameters, an S-shaped speed curve is generated. The opening change sequence corresponding to the S-shaped speed curve is converted into a motor rotation angle sequence and a stepper motor pulse frequency. The pulse frequency increment at adjacent control moments is limited according to the upper limit of the pulse frequency change. The PWM drive command is generated based on the opening adjustment direction. The stepper motor pulse frequency and the PWM drive command are sent to the valve controller. The target execution opening, the actual opening, the water hammer risk constraint, the S-shaped speed curve parameters and the execution deviation are written into the valve execution status.

[0011] Optionally, step five specifically includes: The edge gateway sends communication detection data to the cloud platform according to a preset detection cycle, and generates cloud communication status based on the confirmation results from the cloud platform. When the number of consecutive times no confirmation result is received reaches a preset threshold, or when the time interval between the reception time of the last confirmation result and the current control cycle reaches a preset time threshold, the cloud communication status is determined to meet the preset communication timeout condition. After the cloud communication status meets the preset communication timeout conditions, the validity of the pipeline capacity verification results in the local buffer is checked, and the pipeline capacity verification result that passes the verification and is closest to the current control cycle is determined as the most recent valid pipeline capacity verification result. The most recent effective pipeline capacity verification results are matched with the irrigation status sequence of the current control cycle, and the emergency target opening degree is determined according to the edge emergency control rules. Among them, for electric valves that are in the suspended waiting state or whose pipeline hydraulic state does not meet the water supply conditions, the emergency target opening degree is set to the current opening degree or the closed opening degree. Valve control commands are generated based on the emergency target opening degree, water hammer risk constraints, and valve execution status, and then sent to the valve controller.

[0012] Optionally, step six specifically includes: The remaining deviation of the opening degree, the direction of microstep adjustment, and the pressure fluctuation amplitude are calculated based on the valve opening and closing constraint parameters, valve execution status, and pipeline hydraulic status. When the absolute value of the remaining deviation of the opening degree is not greater than the preset opening dead zone threshold, or the time interval between the last action time and the current control cycle is less than the preset action interval threshold, or the pressure fluctuation amplitude is greater than the preset pressure stability threshold, no restricted microstep adjustment command is generated. When the absolute value of the remaining deviation of the opening is greater than the preset opening dead zone threshold, the time interval is not less than the preset action interval threshold, and the pressure fluctuation amplitude is not greater than the preset pressure stability threshold, the upper limit of the single micro-step opening is determined based on the water hammer risk constraint, the upper limit of the in-place buffer opening is determined based on the end buffer opening interval, and the minimum value among the upper limit of the single micro-step opening, the upper limit of the in-place buffer opening, and the absolute value of the remaining deviation of the opening is determined as the micro-step opening. When the current opening degree is within the end buffer opening degree range, the opening degree of this micro-step is limited to the upper limit of the buffer opening degree; when the historical opening degree deviation is greater than the preset deviation threshold, the opening degree of this micro-step is corrected according to the ratio of the historical opening degree deviation to the preset deviation threshold. Based on the conversion relationship between the microstep opening, microstep adjustment direction, stepper motor step angle, reduction ratio, and valve opening, the number of pulses is generated, and the microstep pulse frequency is generated according to the upper limit of pulse frequency change. The microstep adjustment direction, microstep opening, number of pulses, microstep pulse frequency, and control cycle identifier are combined to generate a restricted microstep adjustment command.

[0013] Optionally, step seven specifically includes: Irrigation execution feedback is generated based on valve control commands, restricted micro-step adjustment commands, pre-irrigation irrigation state sequence, and post-irrigation state changes. The prediction bias is generated based on the difference between the root layer humidity change and the predicted root layer humidity profile, and the parameters of the root layer hydrodynamic mixing prediction model are updated according to the prediction bias. The pipeline response deviation is generated based on actual flow rate changes, pressure changes, and water hammer risk constraints, and the pipeline capacity verification rules are updated based on the pipeline response deviation. The valve execution deviation is generated based on the execution results of the target execution opening degree, the actual actual opening degree, and the restricted microstep adjustment command, and the valve opening and closing constraint parameters are updated according to the valve execution deviation. When the valve control command is a valve control command generated according to the edge emergency control rule, the edge emergency control rule is updated according to the valve control command during the network outage and the changes in the state after irrigation, and the updated root layer hydrodynamic mixing prediction model parameters, network capacity verification rules, valve opening and closing constraint parameters and edge emergency control rules are written into the next control cycle.

[0014] The beneficial effects of this invention are: This invention solves the problem of misjudging root zone water shortage caused by relying on single-point humidity or pure data models in existing irrigation systems. It inputs irrigation state sequences into a root zone hydrodynamic hybrid prediction model, generates a water transport baseline trajectory using the Richards equation, and generates residual memory features using the RWKV-TS network combined with physical residual consistency. Constraint correction is then performed through root zone stress gating. This enables water demand assessment to reflect soil moisture transport patterns, historical irrigation deviations, and the true water shortage state in the root zone. Furthermore, it determines water demand priority based on the predicted humidity profile and water demand rate curve of the root zone, and generates a network capacity verification result including target opening degree of electric valves and water hammer risk constraints, based on the network hydraulic state. This addresses the issue of multiple valves operating simultaneously. The system addresses issues such as excessive water supply capacity, insufficient terminal pressure, and water hammer caused by untimely responses. By compensating and limiting the target opening based on the pipeline capacity verification results, valve flow characteristics, and valve execution status, and dynamically adjusting the S-shaped speed curve according to water hammer risk constraints and valve opening and closing constraints, the system solves the problems of hard start and stop of valves, overshoot of opening, and accumulation of execution deviations. By incorporating the most recent effective pipeline capacity verification results into emergency control for network outages, and combining restricted micro-step adjustment and irrigation execution feedback to update model parameters and control rules, the system can maintain consistent irrigation control even in scenarios with weak networks, pipeline fluctuations, and valve aging. This has practical significance for improving the water-saving performance, stability, and equipment lifespan of large-scale intelligent irrigation for farmland. Attached Figure Description

[0015] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a remote control and intelligent adjustment method based on an intelligent irrigation electric valve proposed in this invention; Figure 2 This is a schematic diagram of a remote control and intelligent adjustment method based on an intelligent irrigation electric valve proposed in this invention; Figure 3 This is a data flow diagram of the root layer hydrodynamic mixing prediction model in a remote control and intelligent regulation method based on an intelligent irrigation electric valve proposed in this invention. Detailed Implementation

[0016] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0017] refer to Figures 1-3 A remote control and intelligent adjustment method based on an intelligent irrigation electric valve includes the following steps: Step 1: Access the irrigation environment status, pipeline hydraulic status, and valve execution status, and generate an irrigation status sequence through edge gateway processing; Step 2: Input the irrigation state sequence into the root zone hydrodynamic mixed prediction model, generate the water transport baseline trajectory based on soil hydraulic parameters through the Richards equation, generate residual memory features through the RWKV-TS network based on the consistency of historical irrigation state and physical residual, and perform constraint correction according to the root zone water shortage state by root zone stress gating, and generate the root zone predicted humidity profile and water demand rate curve. Step 3: Determine water demand priority based on the predicted humidity profile and water demand rate curve of the root layer, and generate network capacity verification results including target opening degree of electric valves and water hammer risk constraints in combination with the hydraulic state of the pipeline network; Step 4: Based on the pipeline capacity verification results, valve flow characteristics and valve execution status, compensate and limit the target opening of the electric valve, generate valve opening and closing constraint parameters, and dynamically adjust the S-shaped speed curve according to the water hammer risk constraint and valve opening and closing constraint parameters to generate stepper motor pulse frequency and PWM drive command. Step 5: After the cloud communication status meets the preset communication timeout conditions, read the most recent effective pipeline capacity verification result and irrigation status sequence, and generate valve control commands according to the edge emergency control rules; Step 6: Generate restricted micro-step adjustment commands based on valve opening and closing constraint parameters, valve execution status, and network hydraulic status; Step 7: Generate irrigation execution feedback based on valve control commands, restricted microstep adjustment commands, and post-irrigation state changes. Update the root layer hydrodynamic mixing prediction model parameters, pipeline capacity verification rules, valve opening and closing constraint parameters, and edge emergency control rules based on the irrigation execution feedback.

[0018] In this embodiment, step one specifically includes: Read the irrigation control zone configuration table, which includes the irrigation control zone number, pipeline branch number, and electric valve number, and establish the binding relationship between the irrigation control zone, pipeline branch, and electric valve; The system accesses the irrigation environment status via sensor interface, the hydraulic status of the pipeline network via pipeline network monitoring interface, and the valve execution status via valve controller communication interface. Perform status format verification and data verification on the accessed status data. Write the status data that fails the verification into the invalid status record, and match the field encoding of the status data that passes the verification according to the device number and register address. Extract the status value, collection time and device number based on the field encoding matching results, and convert the status value into a unified dimension according to the preset dimension conversion table to generate environmental status field, hydraulic status field and execution status field; Based on the edge gateway system clock, the time offset between the acquisition time of each status field and the system clock is read, NTP time correction is performed on the acquisition time, and the corrected acquisition time is mapped to the control cycle identifier according to the preset control cycle. Query the binding relationship based on the equipment number, and write the environmental status field, hydraulic status field, and execution status field under the corresponding irrigation control area number and pipeline branch number. Write an invalid binding mark for status fields for which no binding relationship is found. After the binding relationship is written, state cleaning is performed on the state fields. State cleaning includes removing state fields with invalid binding tags, removing state fields that exceed the preset value range, removing state fields whose change amplitude in adjacent control cycles is greater than the preset mutation threshold, and writing missing tags on state fields that are not received in the current control cycle. The environmental status field, hydraulic status field, and execution status field that have completed status cleaning are matched according to the irrigation control area number, pipeline branch number, and control cycle identifier, and the successfully matched fields are combined into a status combination record within the same control cycle; The state combination records are arranged in chronological order according to the control cycle identifier, and the state combination records belonging to the same irrigation control area within a continuous control cycle are combined to generate an irrigation state sequence.

[0019] In this embodiment, step two specifically includes: The irrigation state sequence is input into the root zone hydrodynamic mixed prediction model, which includes a Richards equation solution layer, an RWKV-TS network, and root zone stress gating. The root layer humidity status, irrigation input status, evapotranspiration drive status and valve control results of the same irrigation control plot within a continuous control cycle are read from the irrigation status sequence. The irrigation input status, evapotranspiration drive status and valve control results are combined into historical irrigation status. Soil moisture characteristic curves and unsaturated hydraulic conductivity curves are established based on soil hydraulic parameters. The root layer is divided into multiple water transport calculation layers according to soil depth and crop root distribution. The Richards equation solution layer employs a finite volume discretization method, dividing the root system layer into multiple water transport calculation layers. Each water transport calculation layer serves as a water balance control volume. Irrigation input states are converted into upper boundary infiltration flux, evapotranspiration-driven states are converted into root uptake terms for each water transport calculation layer according to root distribution weights, and deep drainage states are converted into lower boundary drainage flux. Interlayer water flux is calculated based on the water potential difference between adjacent water transport calculation layers. An implicit iterative method is used, with the water content of each water transport calculation layer in the next control cycle as the solution. Water balance residuals are established based on infiltration flux, drainage flux, root uptake terms, interlayer water flux, and water content changes. When the water balance residual exceeds a preset residual threshold, the infiltration flux or root uptake terms are corrected according to the direction of the water balance residual, and the solution is iterated again until the water balance residual meets the preset residual threshold, generating a water transport baseline trajectory. The difference between the water transport baseline trajectory and the root layer humidity status of the corresponding control period is calculated to generate physical residuals. The consistency of physical residuals is generated based on the sign maintenance state, direction of change and magnitude of change of physical residuals within continuous control periods. Specifically, when physical residuals accumulate continuously in the same direction, a residual consistency state is generated; when the physical residuals reverse their signs, a residual reversal state is generated; and when the magnitude of change of physical residuals exceeds the preset residual fluctuation threshold, a residual mutation state is generated. Historical irrigation status, water transport baseline trajectory, and physical residual consistency are input into the RWKV-TS network. The RWKV-TS network includes a temporal mixing layer, a channel mixing layer, a recursive state cache layer, and a residual output layer. The temporal mixing layer maps the historical irrigation status and physical residual corresponding to the current control cycle into the current state embedding. Then, it performs key mapping, value mapping, and gating mapping on the current state embedding to obtain the key mapping result, value mapping result, and gating mapping result, respectively. The key mapping result is normalized to generate a key vector, the value mapping result is linearly projected to generate a value vector, and the gating mapping result is activated by sigmoid gating to generate a gating vector. The recursive state cache layer generates exponentially decaying weights according to the control cycle interval and the learnable time constant. It uses the exponentially decaying weights to decay the historical key vector and historical value vector cached in the previous control cycle and merges them with the key vector and value vector of the current control cycle to generate an updated recursive state cache. The temporal mixing layer weights the updated recursive state cache and the current value vector according to the gating vector and outputs the result to generate a temporal memory state. The channel fusion layer performs channel fusion on the time memory state, the water transport baseline trajectory, and the current state embedding to generate a mixed state of water change; the recursive state cache layer adjusts the exponential decay weight according to the physical residual consistency, increasing the retention of historical residuals when the residuals are consistently consistent, decreasing the retention of historical residuals when the residuals are reversed, and shielding the physical residual input of the corresponding control period when the residuals change abruptly; the residual output layer generates the residual memory features corresponding to each water transport calculation layer based on the mixed state of water change. The root zone water shortage state is determined based on the water transport baseline trajectory, root zone moisture status, and target crop water content. Root zone stress gating generates a correction gating coefficient based on the root zone water shortage state. When the root zone moisture status is lower than the target crop water content and the number of consecutive water shortage control cycles reaches a preset water shortage cycle threshold, the root zone water shortage state is defined as a continuous water shortage state. Root zone stress gating multiplies the correction component in the residual memory feature pointing towards the increasing water content direction by the first gating coefficient and sets the correction component pointing towards the decreasing water content direction to zero. When the root zone moisture status... When the difference between the current state and the target moisture content of the crop falls within the preset target proximity range, the root zone water shortage state is determined as the target proximity state, and the root zone stress gating multiplies the correction component in the residual memory feature by the second gating coefficient; when the root zone moisture state is higher than the target moisture content of the crop, the root zone water shortage state is determined as the over-wet state, and the root zone stress gating sets the correction component in the residual memory feature pointing to the direction of increasing moisture content to zero, while retaining the correction component pointing to the direction of decreasing moisture content. Here, the first gating coefficient is greater than one, and the second gating coefficient is greater than zero and less than one. The correction component pointing towards the direction of increasing water content in the residual memory features after root zone stress gating is used as the single-layer water demand correction for the corresponding water transport calculation layer. Field capacity limit treatment is applied to the single-layer water demand correction, which ensures that the single-layer water demand correction for any water transport calculation layer does not exceed the single-layer allowable water replenishment limit for that water transport calculation layer. The single-layer allowable water replenishment limit is determined by the difference between the field capacity of the water transport calculation layer and the current water content converted from the root zone humidity status. When the single-layer water demand correction exceeds the single-layer allowable water replenishment limit, the single-layer water demand correction is truncated to the single-layer allowable water replenishment limit, and the correction gating coefficient of root zone stress gating is updated based on the truncated single-layer water demand correction. The residual memory features after root zone stress gating and field water holding capacity limiting are superimposed onto the water transport baseline trajectory to generate a root zone predicted humidity profile. Based on the difference between the root zone predicted humidity profile and the target crop water content, water demand rate curves are generated in the order of control cycles.

[0020] In this embodiment, step three specifically includes: Read the root zone predicted humidity profile, compare the predicted humidity of each water transport calculation layer with the target water content of the crop layer by layer, mark the water transport calculation layer with the predicted humidity below the lower limit of the target water content of the crop as water-deficient layer, and generate the root zone water demand record based on the number of water-deficient layers, the cumulative water deficit of the water-deficient layers and the cumulative water demand of the water demand rate curve in the current control period. Water demand priority is determined based on root zone water demand records. When the water-deficient layer includes the water transport calculation layer corresponding to the preset crop taproot zone, or when the cumulative water deficit of the water-deficient layer is not less than the first water deficit threshold, the corresponding irrigation control cell is written into the first priority. When the cumulative water deficit of the water-deficient layer is less than the first water deficit threshold and not less than the second water deficit threshold, and the water demand rate curve is on an upward trend, the corresponding irrigation control cell is written into the second priority. Other irrigation control cells with water-deficient layers are written into the third priority. Within the same priority, water demand is sorted from largest to smallest according to the cumulative water demand in the current control cycle; when the cumulative water demand is the same, it is sorted from lowest to highest according to the predicted humidity of the water transport calculation layer corresponding to the pre-set crop main root zone; when the predicted humidity is the same, it is sorted from smallest to largest according to the pressure margin of the water supply path, thus generating a water demand queue sequence. Read the hydraulic status of the pipeline network and determine the water supply path of each irrigation control area based on the binding relationship between the irrigation control area, pipeline branches and electric valves; wherein, the pressure margin of the water supply path is determined by the difference between the current pressure status of the water supply path and the lower limit of the end water supply pressure, and the available water supply capacity is obtained by deducting the allocated flow from the current water supply capacity of the pump station; According to the water demand queuing sequence, capacity constraint allocation is performed on each irrigation control cell sequentially. The water demand flow corresponding to the water demand rate curve in the current control cycle is used as the candidate allocation flow. When the candidate allocation flow is not greater than the available water supply capacity and the pressure margin of the water supply path is not less than the preset pressure margin threshold, the candidate allocation flow is determined as the allocation flow. When the candidate allocation flow is greater than the available water supply capacity, but the available water supply capacity is not less than the preset minimum stable water supply flow, and the corresponding irrigation control cell belongs to the first priority, the available water supply capacity is determined as the allocation flow, and the unallocated water demand is written into the next control cycle. When the above conditions are not met, the corresponding irrigation control cell is put into a suspended waiting state. The target opening degree of the electric valve is determined based on the allocated flow rate, the current pipeline pressure status, and the valve flow characteristics of the corresponding electric valve. When there is an opening degree record in the valve flow characteristics that matches the allocated flow rate, the opening degree record is written into the target opening degree of the electric valve. When there is no matching opening degree record, linear interpolation is performed based on the adjacent opening degree record, and the interpolated opening degree is written into the target opening degree of the electric valve. The target opening change is determined based on the difference between the target opening of the electric valve and its current opening. The expected flow change corresponding to the target opening change is determined based on the valve's flow characteristics. A water hammer risk constraint is then generated by combining this with the water supply path pressure margin. Specifically, a first water hammer risk constraint is generated when the expected flow change is not less than a first flow change threshold, or when the water supply path pressure margin is less than a first pressure margin threshold. A second water hammer risk constraint is generated when the expected flow change is less than the first flow change threshold but not less than a second flow change threshold. A third water hammer risk constraint is generated in all other cases. The pipeline capacity verification result is generated by combining the water demand queuing sequence, allocated flow rate, target opening degree of electric valves, hanging waiting status, and water hammer risk constraints.

[0021] In this embodiment, step four specifically includes: Read the allocated flow rate, electric valve target opening degree, and water hammer risk constraint quantity from the pipeline capacity verification results. Also read the valve flow characteristics and valve execution status. The valve flow characteristics are the correspondence between the electric valve opening degree, the pressure difference before and after the valve, and the through flow rate. The valve execution status includes the current opening degree, the target opening degree of the previous control cycle, the actual opening degree of the previous control cycle, the arrival status, and the historical opening degree deviation. The valve flow sensitivity is obtained by determining the target opening range of the electric valve based on the valve flow characteristics and calculating the flow change corresponding to a unit opening change within this range. Read the target opening degree and actual opening degree within the continuous control cycle, and calculate the difference between them. When the direction of the difference remains consistent within the continuous control cycle, write the average difference value into the historical opening degree deviation, and multiply the absolute value of the historical opening degree deviation by the preset compensation coefficient to generate the opening degree compensation amount. When the historical opening degree deviation indicates that the actual opening degree is less than the target opening degree, add the opening degree compensation amount to the target opening degree of the electric valve. When the historical opening degree deviation indicates that the actual opening degree is greater than the target opening degree, deduct the opening degree compensation amount from the target opening degree of the electric valve. When the positioning status is abnormal positioning or stalled, no opening degree compensation is performed, and the current opening degree is used as the starting opening degree for this adjustment. Calculate the expected flow rate corresponding to the compensated target opening based on the valve flow characteristics; when the expected flow rate is greater than the allocated flow rate in the pipeline capacity verification result, find the maximum allowable opening rate that does not exceed the allocated flow rate by going against the valve flow characteristics, and lower the compensated target opening rate to the maximum allowable opening rate to complete the allocation flow rate limiting; The expected flow change is calculated based on the current flow rate corresponding to the current opening degree and the expected flow rate corresponding to the compensated target opening degree. The allowable flow change boundary is determined based on the water hammer risk constraint and the pressure margin of the water supply path. When the expected flow change is greater than the allowable flow change boundary, the compensated target opening degree is reversed along the opening degree adjustment direction until the expected flow change is not greater than the allowable flow change boundary, thus completing the water hammer risk limit and obtaining the target execution opening degree. The valve opening and closing constraint parameters are generated by combining the target opening degree, the valve opening and closing constraint parameters, the valve flow sensitivity, the allowable flow change boundary, and the water hammer risk constraint. The valve opening and closing direction is determined based on the relationship between the target opening degree and the current opening degree. The water hammer risk constraint, single opening adjustment, valve flow sensitivity, and historical opening deviation are used as inputs for curve generation. Adjacent parameter points are selected from the preset S-shaped speed curve parameter calibration table indexed by the water hammer risk constraint, single opening adjustment, valve flow sensitivity, and historical opening deviation. Interpolation is then used to generate the speed limit, acceleration segment duration, constant speed segment duration, deceleration segment duration, end buffer opening range, and pulse frequency change limit for the current control cycle. Based on the speed limit, acceleration phase duration, constant speed phase duration, deceleration phase duration, and end buffer opening range, an S-shaped speed curve for the current control cycle is generated. The S-shaped speed curve starts at the current opening and ends at the target execution opening, and generates an opening change sequence according to the acceleration phase, constant speed phase, deceleration phase, and end buffer phase. The valve opening change sequence is converted into a motor rotation angle sequence, and the number of stepper motor pulses at each control moment is calculated based on the conversion relationship between the stepper motor step angle, reduction ratio and valve opening. The stepper motor pulse frequency is generated based on the difference in the number of stepper motor pulses between adjacent control moments, and the pulse frequency increment between adjacent control moments is limited according to the upper limit of the pulse frequency change. The PWM drive command is generated according to the opening adjustment direction. The stepper motor pulse frequency and the PWM drive command are sent to the valve controller. The target execution opening, the actual opening, the water hammer risk constraint, the S-shaped speed curve parameters and the execution deviation are written into the valve execution status.

[0022] In this embodiment, step five specifically includes: The edge gateway sends communication detection data to the cloud platform according to a preset detection cycle, and generates cloud communication status based on the confirmation results from the cloud platform. The cloud communication status is generated based on the reception status of the confirmation result. The cloud communication status includes normal communication status and communication timeout status. When the number of consecutive times no confirmation result is received reaches a preset number threshold, or the time interval between the reception time of the last confirmation result and the current control cycle reaches a preset time threshold, the cloud communication status is determined to meet the preset communication timeout condition. Once the cloud communication status meets the preset communication timeout conditions, stop using unconfirmed control commands from the cloud platform and read the network capacity verification results from the local cache. The validity of the read pipeline capacity verification results is checked. The validity check includes verification mark verification, cache period verification and pipeline branch matching verification. The pipeline capacity verification result that passes the verification and is closest to the current control period is determined as the most recent valid pipeline capacity verification result. Read the irrigation status sequence of the current control cycle, and match the irrigation status sequence with the most recent effective pipeline capacity verification result according to the irrigation control area number, pipeline branch number and electric valve number; The edge emergency control rules are invoked for judgment. The edge emergency control rules are local valve control judgment rules pre-set in the edge gateway. They are used to determine the emergency target opening degree based on the latest effective pipeline capacity verification result and irrigation status sequence after the cloud communication status meets the preset communication timeout condition. When the hydraulic condition of the pipeline network meets the water supply conditions, and the corresponding irrigation control area in the most recent effective pipeline network capacity verification result is not in a suspended waiting state, the target opening degree of the electric valve in the most recent effective pipeline network capacity verification result is written into the emergency target opening degree. When the hydraulic condition of the pipeline network does not meet the water supply conditions, or when the corresponding irrigation control area is in a suspended waiting state in the most recent effective pipeline network capacity verification results, the emergency target opening degree will be set to the current opening degree or closed opening degree. When the root layer moisture state in the irrigation state sequence is lower than the crop target water content state, and the emergency target opening degree is greater than the electric valve target opening degree in the most recent effective pipeline capacity verification result, the emergency target opening degree will be limited to the electric valve target opening degree in the most recent effective pipeline capacity verification result. When the root zone moisture state in the irrigation state sequence is higher than the target crop water content state, the emergency target opening degree is set to the closed opening degree; Valve control commands are generated based on the emergency target opening degree, current opening degree, water hammer risk constraint, and valve execution status. The valve control commands include the electric valve number, control cycle identifier, opening and closing direction, emergency target opening degree, and water hammer risk constraint. The valve control command is sent to the valve controller, and the cloud communication status, the latest effective pipeline capacity verification result, the irrigation status sequence, the judgment result of the edge emergency control rule, the emergency target opening degree and the valve control command are written into the local control record.

[0023] In this embodiment, step six specifically includes: The system reads valve opening and closing constraint parameters, valve execution status, and pipeline hydraulic status. Valve opening and closing constraint parameters include target execution opening degree, end buffer opening degree range, pulse frequency variation upper limit, and water hammer risk constraint amount. Valve execution status includes current opening degree, actual arrival opening degree, last action time, and historical opening degree deviation. It generates a residual opening degree deviation based on the difference between the target execution opening degree and the current opening degree, and determines the micro-step adjustment direction based on the sign of the residual opening degree deviation. It generates pressure fluctuation amplitude based on the difference between the pipeline pressure in the current control cycle and the pipeline pressure in the previous control cycle. When the absolute value of the residual opening degree deviation is not greater than the preset opening degree dead zone threshold, or the time interval between the last action time and the current control cycle is less than the preset action interval threshold, or the pressure fluctuation amplitude is greater than the preset pressure stability threshold, no restricted micro-step adjustment command is generated. When the absolute value of the remaining deviation of the opening is greater than the preset opening dead zone threshold, the time interval between the last action time and the current control cycle is not less than the preset action interval threshold, and the pressure fluctuation amplitude is not greater than the preset pressure stability threshold, the micro-step adjustment calculation is initiated. The upper limit of the single micro-step opening is determined based on the water hammer risk constraint, and the upper limit of the position buffer opening is determined based on the end buffer opening range. The minimum value among the upper limit of the single micro-step opening, the upper limit of the position buffer opening, and the absolute value of the remaining deviation of the opening is determined as the opening of this micro-step. When the current opening is within the end buffer opening range, the opening of this micro-step is limited to the upper limit of the position buffer opening. When the historical opening deviation is greater than the preset deviation threshold, the reduction ratio is determined according to the ratio of the historical opening deviation to the preset deviation threshold, and the opening of this micro-step is corrected according to the reduction ratio. Based on the conversion relationship between the current microstep opening, microstep adjustment direction, stepper motor step angle, reduction ratio, and valve opening, the number of pulses corresponding to this microstep adjustment is calculated, and the microstep pulse frequency is generated according to the upper limit of pulse frequency change. The microstep adjustment direction, microstep opening, number of pulses, microstep pulse frequency, and control cycle identifier are combined to generate a restricted microstep adjustment command. The restricted microstep adjustment command is sent to the valve controller, and the current microstep opening, microstep adjustment direction, execution time, and actual opening after execution are written into the valve execution status.

[0024] In this embodiment, step seven specifically includes: Read valve control commands, restricted microstep adjustment commands, pre-irrigation state sequence and post-irrigation state changes, and match them according to irrigation control area number, pipeline branch number and electric valve number to generate irrigation execution feedback; The prediction bias is generated based on the difference between the root zone humidity change in the irrigation execution feedback and the predicted root zone humidity profile. The parameters of the root zone hydrodynamic mixing prediction model are updated based on the prediction bias. The parameters of the root zone hydrodynamic mixing prediction model include soil hydraulic parameters, time decay parameters of the RWKV-TS network and the modified gating coefficient of root zone stress gating. The pipeline response deviation is generated based on the actual flow rate changes, pressure changes, and water hammer risk constraints in the irrigation execution feedback. The available water supply capacity, pressure margin threshold, and water hammer risk constraint generation rules in the pipeline capacity verification rules are updated based on the pipeline response deviation. Based on the execution results of the target opening degree, actual opening degree, and restricted microstep adjustment command in the irrigation execution feedback, valve execution deviation is generated, and the opening degree compensation amount, end buffer opening degree range, and pulse frequency change upper limit in the valve opening and closing constraint parameters are updated according to the valve execution deviation. When the valve control command is a valve control command generated according to the edge emergency control rule, the edge emergency control rule is updated according to the valve control command during the network outage and the changes in the state after irrigation, and the updated root layer hydrodynamic mixing prediction model parameters, network capacity verification rules, valve opening and closing constraint parameters and edge emergency control rules are written into the next control cycle.

[0025] Example 1: To verify the feasibility of this invention in practice, it was applied to a remote control scenario for intelligent irrigation electric valves in a facility agriculture demonstration park. This demonstration park includes 6 irrigation control zones, 2 pipeline branches, and 18 electric valves. The crops are tomatoes and cucumbers. The root system is divided into multiple water transport calculation layers at 10cm, 20cm, and 40cm depths. The edge gateway is deployed in the pump room control cabinet. The preset control cycle is 15 minutes, the preset opening dead zone threshold is 1.0%, the preset action interval threshold is 300 seconds, the preset pressure stability threshold is 0.08 MPa, the preset residual threshold is 0.015, the preset pressure margin threshold is 0.05 MPa, and the preset minimum stable water supply flow rate is 1.2 m³ / h.

[0026] During operation, the edge gateway accesses irrigation environment status, network hydraulic status, and valve execution status to generate an irrigation status sequence. The root zone hydrodynamic hybrid prediction model first calculates the baseline trajectory of water transport using the Richards equations solution layer. Then, the RWKV-TS network generates residual memory features based on historical irrigation status and physical residual consistency. Root zone stress gating is constrained and corrected according to the root zone water shortage state, and field capacity limiting is applied to the single-layer water demand correction. The system outputs a predicted humidity profile and water demand rate curve for the root zone. Subsequently, the system generates network capacity verification results based on water demand priority, water supply path pressure margin, and available water supply capacity, including allocated flow rate, target opening of electric valves, pending status, and water hammer risk constraints. During the valve execution phase, the target opening is compensated and limited based on valve flow characteristics and valve execution status. The S-shaped velocity curve is dynamically adjusted using water hammer risk constraints and valve opening / closing constraint parameters to generate stepper motor pulse frequency and PWM drive commands.

[0027] To compare the results, Method A was set as traditional humidity threshold control, which uses a single-point soil moisture level below the threshold to open the valve; Method B was LSTM water demand prediction with fixed opening control; Method C was the unmodified model, which uses Richards equations with ordinary RWKV-TS prediction, but does not include physical residual consistency, root zone stress gating, field capacity limiting, and dynamic S-shaped velocity curve; The test period was 30 consecutive days, including sunny days, consecutive cloudy days, and one cloud communication interruption scenario.

[0028] Table 1. Comparison of different methods in the control of electric valves for intelligent irrigation

[0029] As shown in Table 1, Method A, relying solely on humidity thresholds, is prone to misjudging surface moisture while root zone water deficiency, with an average root zone humidity prediction error of 12.8%. Furthermore, it involves frequent valve actuation, indicating a tendency for frequent opening and closing near critical humidity levels. While Method B incorporates LSTM water demand prediction, it lacks the water balance constraint of the Richards equation and the network capacity verification results to constrain the target valve opening, resulting in a still high number of instances of insufficient end-point pressure and high-risk water hammer triggering. Method C incorporates physical and temporal predictions, but lacks physical residual consistency, root zone stress gating, and field capacity limiting. Prediction errors accumulate after consecutive cloudy days, and the fixed S-shaped velocity curve cannot dynamically adjust with water hammer risk constraints.

[0030] The average error of root zone moisture prediction in this invention was reduced to 3.7%, indicating that the water transport baseline trajectory provided by the Richards equation solution layer, the residual memory characteristics formed by the RWKV-TS network, and the constraint correction of root zone stress gating can more stably reflect changes in root zone moisture. The number of times the terminal pressure was insufficient decreased from 16 times in Method A to 2 times, indicating that the problem of multiple valves competing for water was significantly alleviated after generating network capacity verification results based on water demand priority and network hydraulic status. The number of high-risk water hammer triggers decreased from 11 times to 1 time, indicating that the valve opening and closing process was smoother after the water hammer risk constraint was incorporated into the valve opening and closing constraint parameters and the dynamic adjustment of the S-shaped velocity curve. The average number of valve actions per day decreased from 42 times to 15 times, indicating that the restricted micro-step adjustment command reduced ineffective fine-tuning through opening dead zone, action interval, and pressure stability conditions.

[0031] As can be seen from this scenario, this invention does not simply improve prediction accuracy, but rather integrates the predicted humidity profile of the root zone, the water demand rate curve, the pipeline capacity verification results, valve opening and closing constraint parameters, and irrigation execution feedback into a continuous control link, making water demand judgment, pipeline scheduling, and valve execution mutually constrained. This method can reduce water consumption, reduce valve impact and network outage malfunctions in large-scale irrigation scenarios, and improve the root zone moisture compliance rate, demonstrating significant engineering application value.

[0032] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for remote control and intelligent adjustment based on intelligent irrigation electric valve, characterized in that, Includes the following steps: Step 1: Access the irrigation environment status, pipeline hydraulic status, and valve execution status, and generate an irrigation status sequence through edge gateway processing; Step 2: Input the irrigation state sequence into the root zone hydrodynamic mixed prediction model, generate the water transport baseline trajectory based on soil hydraulic parameters through the Richards equation, generate residual memory features through the RWKV-TS network based on the consistency of historical irrigation state and physical residual, and perform constraint correction according to the root zone water shortage state by root zone stress gating, and generate the root zone predicted humidity profile and water demand rate curve. Step 3: Determine water demand priority based on the predicted humidity profile and water demand rate curve of the root layer, and generate network capacity verification results including target opening degree of electric valves and water hammer risk constraints in combination with the hydraulic state of the pipeline network; Step 4: Based on the pipeline capacity verification results, valve flow characteristics and valve execution status, compensate and limit the target opening of the electric valve, generate valve opening and closing constraint parameters, and dynamically adjust the S-shaped speed curve according to the water hammer risk constraint and valve opening and closing constraint parameters to generate stepper motor pulse frequency and PWM drive command. Step 5: After the cloud communication status meets the preset communication timeout conditions, read the most recent effective pipeline capacity verification result and irrigation status sequence, and generate valve control commands according to the edge emergency control rules; Step 6: Generate restricted micro-step adjustment commands based on valve opening and closing constraint parameters, valve execution status, and network hydraulic status; Step 7: Generate irrigation execution feedback based on valve control commands, restricted microstep adjustment commands, and post-irrigation state changes. Update the root layer hydrodynamic mixing prediction model parameters, pipeline capacity verification rules, valve opening and closing constraint parameters, and edge emergency control rules based on the irrigation execution feedback.

2. The remote control and intelligent adjustment method based on an intelligent irrigation electric valve according to claim 1, characterized in that, Step one specifically includes: Establish the binding relationship between irrigation control zones, pipeline branches, and electric valves based on the irrigation control zone configuration table; The system accesses the irrigation environment status via sensor interface, the hydraulic status of the pipeline network via pipeline network monitoring interface, and the valve execution status via valve controller communication interface. Perform field encoding matching, status value extraction, and dimension conversion on irrigation environment status, pipeline hydraulic status, and valve execution status to generate environment status field, hydraulic status field, and execution status field; The acquisition time is corrected using the edge gateway system clock as a reference, and a control cycle identifier is generated according to the preset control cycle. Perform invalid binding removal, value range verification, adjacent control cycle mutation verification, and missing field marking on the environmental status field, hydraulic status field, and execution status field; The fields that have completed status cleaning are matched according to the irrigation control area number, pipeline branch number and control cycle identifier to generate status combination records, and then combined according to the time order of the control cycle identifier to generate an irrigation status sequence.

3. The remote control and intelligent adjustment method based on an intelligent irrigation electric valve according to claim 1, characterized in that, Step two specifically includes: The root zone hydrodynamic mixing prediction model includes a Richards equation solution layer, an RWKV-TS network, and root zone stress gating. Historical irrigation states are formed based on root layer humidity, irrigation input, evapotranspiration and valve control results in the irrigation state sequence. Soil moisture characteristic curves and unsaturated hydraulic conductivity curves are established based on soil hydraulic parameters, and the root layer is divided into multiple water transport calculation layers. The Richards equation solution layer uses a finite volume discretization method, taking each water transport calculation layer as a water balance control volume, and adopts an implicit iteration method, taking the water content of each water transport calculation layer in the next control period as the solution quantity, and calculating the water balance residual based on infiltration flux, drainage flux, root water uptake, interlayer water flux and water content change. When the water balance residual exceeds the preset residual threshold, the infiltration flux or root water absorption term is corrected according to the direction of the water balance residual and the solution is iterated again until the water balance residual meets the preset residual threshold, and a water transport baseline trajectory is generated. Physical residuals are generated based on the difference between the water transport baseline trajectory and the root layer humidity state, and physical residual consistency is generated according to the continuous same-direction accumulation, sign reversal and abrupt change states of physical residuals. The RWKV-TS network processes historical irrigation states, water transport baseline trajectories, and physical residual consistency through a temporal mixing layer, a channel mixing layer, a recursive state buffer layer, and a residual output layer. The temporal mixing layer generates key vectors, value vectors, and gating vectors based on state embedding. The recursive state buffer layer generates exponentially decaying weights based on the control period interval and a learnable time constant, and adjusts the degree of historical residual retention according to physical residual consistency. The residual output layer generates residual memory features. The root zone stress gating generates a modified gating coefficient based on the water shortage state of the root layer. It performs gating processing on the water content increase direction correction component and water content decrease direction correction component in the residual memory characteristics, and performs field capacity limit processing on the single layer water demand correction, generating a predicted humidity profile and water demand rate curve for the root layer.

4. The remote control and intelligent adjustment method based on an intelligent irrigation electric valve according to claim 1, characterized in that, Step three specifically includes: The water-deficient layer is determined based on the water transport calculation layer below the target moisture state limit of the crop in the root layer predicted humidity profile, and the root zone water demand record is generated by combining the cumulative water deficit of the water-deficient layer and the water demand rate curve. Water demand priority is determined based on whether the water-deficient layer includes the preset crop taproot zone, the cumulative water deficit of the water-deficient layer, and the trend of water demand rate curve. A water demand queue is generated based on the cumulative water demand, the predicted humidity of the taproot zone, and the pressure margin of the water supply path. The water supply path is determined based on the binding relationship between irrigation control areas, pipeline branches and electric valves, and the water demand queuing sequence is allocated with capacity constraints based on the pressure margin of the water supply path and the available water supply capacity, generating the allocated flow or the suspended waiting state. The target opening degree of the electric valve is determined based on the allocated flow rate, the current pipeline pressure status, and the valve flow characteristics. The water hammer risk constraint is generated based on the change in the target opening degree, the expected change in flow rate, and the pressure margin of the water supply path. The pipeline capacity verification result is generated by combining the water demand queuing sequence, allocated flow rate, target opening degree of electric valves, hanging waiting status, and water hammer risk constraints.

5. The remote control and intelligent adjustment method based on an intelligent irrigation electric valve according to claim 1, characterized in that, Step four specifically includes: The valve flow sensitivity corresponding to the target opening degree of the electric valve is determined based on the valve flow characteristics, and the opening compensation amount is generated based on the continuous deviation between the target opening degree and the actual opening degree to compensate for the target opening degree of the electric valve. Based on the allocated flow rate, the allocated flow rate is limited to the compensated target opening. Based on the water hammer risk constraint and the water supply path pressure margin, the allowable flow rate change boundary is determined. The water hammer risk is limited to the compensated target opening to obtain the target opening. Valve opening and closing constraint parameters are generated based on the target opening degree, opening degree adjustment direction, single opening degree adjustment amount, valve flow sensitivity, allowable flow change boundary and water hammer risk constraint. Using water hammer risk constraints, single opening adjustment, valve flow sensitivity, and historical opening deviation as indexes, adjacent parameter points are selected in the preset S-shaped speed curve parameter calibration table and interpolated to generate the speed limit, acceleration segment duration, constant speed segment duration, deceleration segment duration, end buffer opening range, and pulse frequency change limit. Based on the above parameters, an S-shaped speed curve is generated. The opening change sequence corresponding to the S-shaped speed curve is converted into a motor rotation angle sequence and a stepper motor pulse frequency. The pulse frequency increment at adjacent control moments is limited according to the upper limit of the pulse frequency change. The PWM drive command is generated based on the opening adjustment direction. The stepper motor pulse frequency and the PWM drive command are sent to the valve controller. The target execution opening, the actual opening, the water hammer risk constraint, the S-shaped speed curve parameters and the execution deviation are written into the valve execution status.

6. The remote control and intelligent adjustment method based on an intelligent irrigation electric valve according to claim 1, characterized in that, Step five specifically includes: The edge gateway sends communication detection data to the cloud platform according to a preset detection cycle, and generates cloud communication status based on the confirmation results from the cloud platform. When the number of consecutive times no confirmation result is received reaches a preset threshold, or when the time interval between the reception time of the last confirmation result and the current control cycle reaches a preset time threshold, the cloud communication status is determined to meet the preset communication timeout condition. After the cloud communication status meets the preset communication timeout conditions, the validity of the pipeline capacity verification results in the local buffer is checked, and the pipeline capacity verification result that passes the verification and is closest to the current control cycle is determined as the most recent valid pipeline capacity verification result. The most recent effective pipeline capacity verification results are matched with the irrigation status sequence of the current control cycle, and the emergency target opening degree is determined according to the edge emergency control rules. Among them, for electric valves that are in the suspended waiting state or whose pipeline hydraulic state does not meet the water supply conditions, the emergency target opening degree is set to the current opening degree or the closed opening degree. Valve control commands are generated based on the emergency target opening degree, water hammer risk constraints, and valve execution status, and then sent to the valve controller.

7. The remote control and intelligent adjustment method based on an intelligent irrigation electric valve according to claim 1, characterized in that, Step six specifically includes: The remaining deviation of the opening degree, the direction of microstep adjustment, and the pressure fluctuation amplitude are calculated based on the valve opening and closing constraint parameters, valve execution status, and pipeline hydraulic status. When the absolute value of the remaining deviation of the opening degree is not greater than the preset opening dead zone threshold, or the time interval between the last action time and the current control cycle is less than the preset action interval threshold, or the pressure fluctuation amplitude is greater than the preset pressure stability threshold, no restricted microstep adjustment command is generated. When the absolute value of the remaining deviation of the opening is greater than the preset opening dead zone threshold, the time interval is not less than the preset action interval threshold, and the pressure fluctuation amplitude is not greater than the preset pressure stability threshold, the upper limit of the single micro-step opening is determined based on the water hammer risk constraint, the upper limit of the in-place buffer opening is determined based on the end buffer opening interval, and the minimum value among the upper limit of the single micro-step opening, the upper limit of the in-place buffer opening, and the absolute value of the remaining deviation of the opening is determined as the micro-step opening. When the current opening degree is within the end buffer opening degree range, the opening degree of this micro-step is limited to the upper limit of the buffer opening degree; when the historical opening degree deviation is greater than the preset deviation threshold, the opening degree of this micro-step is corrected according to the ratio of the historical opening degree deviation to the preset deviation threshold. Based on the conversion relationship between the microstep opening, microstep adjustment direction, stepper motor step angle, reduction ratio, and valve opening, the number of pulses is generated, and the microstep pulse frequency is generated according to the upper limit of pulse frequency change. The microstep adjustment direction, microstep opening, number of pulses, microstep pulse frequency, and control cycle identifier are combined to generate a restricted microstep adjustment command.

8. A remote control and intelligent adjustment method based on an intelligent irrigation electric valve according to claim 1, characterized in that, Step seven specifically includes: Irrigation execution feedback is generated based on valve control commands, restricted micro-step adjustment commands, pre-irrigation irrigation state sequence, and post-irrigation state changes. The prediction bias is generated based on the difference between the root layer humidity change and the predicted root layer humidity profile, and the parameters of the root layer hydrodynamic mixing prediction model are updated according to the prediction bias. The pipeline response deviation is generated based on actual flow rate changes, pressure changes, and water hammer risk constraints, and the pipeline capacity verification rules are updated based on the pipeline response deviation. The valve execution deviation is generated based on the execution results of the target execution opening degree, the actual actual opening degree, and the restricted microstep adjustment command, and the valve opening and closing constraint parameters are updated according to the valve execution deviation. When the valve control command is a valve control command generated according to the edge emergency control rule, the edge emergency control rule is updated according to the valve control command during the network outage and the changes in the state after irrigation, and the updated root layer hydrodynamic mixing prediction model parameters, network capacity verification rules, valve opening and closing constraint parameters and edge emergency control rules are written into the next control cycle.

Images