Moisture-machinery cooperative intelligent management system suitable for multi-crop sloping field

Abstract

The invention relates to a moisture-machinery cooperative intelligent management system suitable for a multi-crop sloping field, and belongs to the technical field of agricultural irrigation automation. The system aims at solving the problems that in slope planting, mechanical operation is not well matched with an irrigation system, and water and fertilizer regulation and control accuracy is insufficient, the system adapts to different slope terrains and crop line spacing through a wheel-track-adjustable self-propelled sprinkling irrigation machine, flow is intelligently adjusted according to data collected by a slope sensor through a slope self-adaptive drip irrigation unit, and the water and fertilizer regulation and control accuracy is improved. And by means of the intelligent sensing and control module, the soil moisture content, fertility and meteorological data are fused, a water-fertilizer ratio and an operation instruction are intelligently decided according to a multi-crop multi-growth-period threshold value, and finally precise collaborative management and efficient utilization of water-fertilizer resources in the slope environment are achieved.

Description

Technical Field

[0001] This invention belongs to the field of agricultural irrigation automation technology, specifically relating to a water-mechanical collaborative intelligent management system suitable for multi-crop slopes. Background Technology

[0002] In hillside agriculture, mechanized irrigation often faces adaptability challenges due to complex terrain and diverse crop types. Traditional sprinkler irrigation equipment is mostly designed for flat terrain and struggles to operate stably on slopes with significant variations. The undulating terrain also causes instability in the equipment's center of gravity, making ordinary wheeled machinery prone to skidding or tipping over, thus limiting its applicability on slopes.

[0003] In terms of irrigation regulation, sloping terrain has a significant impact on water distribution, with uneven water distribution occurring between different slope positions within the same field. Traditional irrigation systems often lack real-time responsiveness to slope factors, failing to dynamically adjust irrigation volume according to slope changes. This frequently leads to insufficient irrigation in the upper slope areas while water accumulates or even runsoff occurs in the lower slope areas, severely impacting crop growth uniformity and water resource utilization efficiency. Furthermore, in integrated water and fertilizer management, due to the lack of comprehensive judgment and decision-making based on multi-source environmental information, fertilizer application amount and timing often rely on manual experience, making it difficult to achieve precise regulation according to different crops and different growth stages, easily resulting in fertilizer waste or localized nutrient deficiencies.

[0004] Therefore, in multi-crop planting environments on slopes, achieving adaptability between mechanized irrigation equipment and terrain, as well as improving the precision of water and fertilizer regulation, has become an urgent problem to be solved in current agricultural irrigation technology. Although some technologies have attempted to alleviate these problems by improving mechanical structures or introducing single-parameter control, a systematic solution is still lacking that can collaboratively achieve multiple goals such as adaptability, uniform irrigation, and precise management of fertilizer and water under complex slope conditions. Summary of the Invention

[0005] One object of the present invention is to solve at least the above-mentioned problems and to provide at least the advantages that will be described later.

[0006] Another objective of this invention is to provide a water-mechanism collaborative intelligent management system suitable for multi-crop slopes. This system can achieve uniform irrigation of the slope by utilizing slope sensing and flow feedback control, and complete precise water and fertilizer ratio and collaborative operation based on multi-source sensor data fusion and decision-making algorithms, thereby improving the efficiency of water and machinery management for multi-crop planting on slopes.

[0007] To achieve these objectives and other advantages of the present invention, a water-mechanism collaborative intelligent management system suitable for multi-crop slopes is provided, including an adjustable wheel track self-propelled sprinkler irrigation machine, a slope adaptive drip irrigation unit, and an intelligent sensing and control module. The adjustable wheel track self-propelled sprinkler irrigation machine includes a wheel track adjustment mechanism, a rubber track walking mechanism, a fertilizer storage tank, and a fault self-diagnosis module. The fault self-diagnosis module monitors the motor speed and pipeline pressure in real time. When the motor speed fluctuates by more than ±5% or the pipeline pressure fluctuates by more than ±0.2MPa, the module automatically stops the machine and pushes the fault code to the intelligent sensing and control module. The slope adaptive drip irrigation unit includes drip irrigation pipelines, slope sensors, and electromagnetic flow regulating valves. The interval of the drip irrigation pipelines is set according to the crop row spacing. The slope sensors collect the slope angle in real time, and the electromagnetic flow regulating valves adjust the water output according to the slope angle. The intelligent sensing and control module includes a soil moisture sensor, a weather station, a soil fertility sensor, and an embedded controller. Soil moisture sensors are buried at a depth of 20cm to 30cm, and soil fertility sensors are buried at a depth of 15cm to 20cm. The embedded controller receives data from soil moisture sensors, weather stations, and soil fertility sensors via a wireless network; The embedded controller presets irrigation thresholds for multiple crops and growth stages. Based on slope sensor data, crop row spacing type, and soil moisture content data, the embedded controller makes decisions and starts adjustable wheel spacing self-propelled sprinkler irrigation machine or slope adaptive drip irrigation unit for irrigation operations. Based on weather station data and soil fertility sensor data, the embedded controller generates fertilization ratio scheme and controls the fertilizer storage tank to perform variable fertilization. The embedded controller supports remote control of irrigation and fertilization amounts via mobile terminal application.

[0008] Preferably, the fault self-diagnosis module of the adjustable wheel track self-propelled sprinkler also integrates a GPS positioning module. The positioning accuracy of the GPS positioning module is within 10m. When the fault self-diagnosis module pushes a fault code, it simultaneously sends the real-time location information of the sprinkler. After receiving the fault code and location information, the embedded controller automatically generates a maintenance instruction containing the location of the faulty equipment and the fault type. The maintenance instruction is pushed to the mobile terminal of the maintenance personnel through a 4G or 5G network.

[0009] Preferably, the embedded controller is also connected to a local data storage module, which records the historical fault codes, maintenance records, and location history information of each adjustable wheel track self-propelled sprinkler. The maintenance instructions also include the three most recent fault records and the most recent maintenance record of the corresponding sprinkler. When the fault self-diagnosis module pushes the same fault code repeatedly, the embedded controller adds a warning indicator to the maintenance instructions. The maintenance records include the name of the repaired part, the maintenance method, the start and end time of the maintenance, and the model of the replaced part.

[0010] Preferably, the embedded controller also integrates a fault mode analysis module; the fault mode analysis module analyzes the correlation between fault codes based on historical fault codes and maintenance record data, using an association rule algorithm. When the fault mode analysis module receives a new fault code, it automatically matches other potentially related system components and generates a preventative checklist. The preventive inspection checklist is pushed to the maintenance personnel's mobile terminal along with the maintenance instructions; The preventative inspection checklist includes the names of the components to be inspected, the inspection methods, and any potential faults that may exist.

[0011] Preferably, the specific implementation of the association rule algorithm in the fault mode analysis module includes: The data preprocessing unit organizes historical fault codes into transaction datasets by device number and time sequence. Each transaction contains all fault codes recorded in a single maintenance event. The frequent itemset mining unit uses the FP-Growth algorithm to scan the transaction dataset and find frequent fault code itemsets with support between 2% and 5%. The rule generation unit generates association rules with a confidence level between 60% and 80% based on frequent itemsets; The rule filtering unit uses the lift index to evaluate the effectiveness of association rules, and only retains association rules with a lift greater than one for the generation of preventive checklists; The rule filtering unit automatically recalculates the support, confidence, and lift of all associated rules every month. When the confidence of a rule drops by more than 20% or the support drops by more than 50%, the rule is automatically removed from the valid rule base.

[0012] Preferably, the embedded controller includes a device decision logic unit; The device decision logic unit receives slope sensor data, crop row spacing type data, and soil moisture content data; The equipment decision logic unit first determines the equipment type based on slope sensor data: When the slope sensor data is less than 10°, the decision is made to start the adjustable wheel track self-propelled sprinkler irrigation machine. When the slope sensor data is greater than or equal to 10° and less than or equal to 25°, the decision is made to activate the slope adaptive drip irrigation unit. Subsequently, the equipment decision logic unit makes irrigation intensity decisions based on soil moisture content data: When the soil moisture content is more than 5 percentage points lower than the threshold for the corresponding crop growth period, if the decision has been made to start the adjustable wheel track self-propelled sprinkler irrigation machine, then control it to perform 130% of the standard irrigation volume. If the decision has been made to activate the slope adaptive drip irrigation unit, then control it to operate for 140% of the standard irrigation duration; The equipment decision logic unit reassesses the soil moisture content data and updates the irrigation intensity control instructions every 30 minutes.

[0013] Preferably, the device decision logic unit is also connected to a data reliability assessment module; The data reliability assessment module monitors the data output status of the slope sensor and soil moisture sensor in real time; When the standard deviation of three consecutive slope sensor readings is greater than 1°, the sensor data is deemed unreliable. When the coefficient of variation of soil moisture sensor data collected over 10 consecutive minutes is greater than 15%, the sensor data is deemed unreliable. When any sensor data is determined to be unreliable, the data reliability assessment module sends a data failure signal to the device decision logic unit. After receiving a data failure signal, the equipment decision logic unit automatically switches to a preset irrigation mode based on crop growth period and weather station data. In the preset irrigation mode, the equipment decision logic unit performs irrigation operations according to the standard irrigation amount for the current crop growth period, and reassesses the reliability status of sensor data every 6 hours.

[0014] Preferably, the embedded controller includes a fertilization decision unit; The fertilization decision unit receives precipitation probability data for the next 24 hours from the meteorological station and nitrogen, phosphorus, and potassium content data from the soil fertility sensor. The fertilization decision unit has built-in threshold ranges for nutrient requirements at different growth stages of different crops; When the content of a certain element in the soil is lower than the lower limit of the threshold required by the corresponding crop at the current growth stage, the fertilization decision unit generates a supplementary formula for that element. When the probability of precipitation in the next 24 hours is greater than 50%, the fertilization decision-making unit will postpone the implementation of the fertilization plan until after the precipitation event ends; The fertilization decision unit generates fertilization control instructions and sends them to the adjustable wheel track self-propelled sprinkler irrigation machine to control the mixing ratio of nitrogen, phosphorus and potassium fertilizers in its fertilizer storage tank, with the mixing accuracy controlled within ±3%. The fertilization decision unit dynamically adjusts the fertilization interval to 3–7 days based on the rate of change of soil fertility sensor data.

[0015] Preferably, the fertilization decision unit is also connected to a fertilization effect feedback and evaluation module; The fertilization effect feedback and evaluation module is activated after each fertilization plan is executed, and continuously monitors the changes in soil fertility sensor data. This module sets the expected rate of increase in nitrogen concentration to a threshold of 0.5–1.0 mg / kg per hour; When the actual monitored rate of increase in nitrogen concentration is lower than the lower limit of the expected threshold, the fertilization effect feedback evaluation module sends a signal of insufficient fertilizer effect to the fertilization decision unit. After receiving a signal of insufficient fertilizer effectiveness, the fertilization decision-making unit automatically initiates a secondary precision topdressing procedure. The secondary precision topdressing process adds 20% to 30% of the original fertilizer amount in the form of liquid fertilizer through an adjustable wheel track self-propelled sprinkler irrigation machine. The fertilization effect feedback and evaluation module continues to monitor for 2 hours after additional fertilization. If the fertilizer effect is still insufficient, it will generate a sensor calibration prompt.

[0016] The present invention has at least the following beneficial effects: First, the system of the present invention improves the passability and stability of the equipment in sloping terrain through the adjustable wheel track structure and track walking mechanism, which can adapt to different crop row spacing requirements and avoid the problems of slippage, tilting or crushing crops that are easy to occur in traditional fixed wheel track equipment in sloping operations, thereby improving the adaptability and safety of mechanical operation.

[0017] Secondly, the system of this invention is based on real-time slope sensing and flow control technology, which realizes dynamic allocation of irrigation water on slopes, overcomes the shortcomings of uneven water distribution on slopes in traditional irrigation methods, effectively reduces runoff loss and irrigation blind spots, and improves water use efficiency and crop growth consistency.

[0018] Third, by integrating multi-source data such as soil moisture, fertility, and meteorology, the system can make precise decisions and controls on the water and fertilizer ratio based on the crop growth stage, avoiding over- or under-fertilization caused by relying on human experience, thereby improving fertilizer utilization while reducing environmental impact.

[0019] Fourth, this invention breaks through the bottleneck of existing technologies being limited to "single function improvement" and for the first time realizes the systematic integration and coordinated operation of "mechanical adaptation (adjustable wheel spacing) + uniform irrigation (dynamic slope control) + water and fertilizer synergy (multi-source data decision-making) + intelligent fault operation and maintenance (association rule analysis)". The core innovation lies in the synergistic effect of multiple modules: First, by using an adjustable wheel-track sprinkler to adapt to different crop row spacings, combined with a slope-adaptive drip irrigation unit responding to slope changes, the system can intelligently switch between modules, overcoming the limitation of traditional equipment that can only adapt to row spacing or slope. This improves irrigation uniformity to over 88% for different slopes (5°–25°) and different crops (such as wheat and citrus). Second, by integrating fault self-diagnosis (real-time monitoring of motor and pipeline pressure), GPS positioning, and intelligent analysis based on association rules, the average fault handling time is significantly reduced from 120 minutes to 25 minutes, far exceeding the effect of single fault alarms or simple location technology. Finally, through a fertilization decision-making mechanism that integrates soil fertility and rainfall probability, combined with secondary topdressing feedback control, the system effectively avoids the problems of "fertilizer loss due to rainfall after soil testing" or "waste due to blind topdressing," increasing fertilizer utilization by 17%. This invention, through efficient multi-module synergy, significantly improves the overall performance of slope agriculture in terms of water and fertilizer management, equipment operation and maintenance, and operational adaptability.

[0020] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Detailed Implementation

[0021] The present invention will now be described in further detail so that those skilled in the art can implement it based on the description.

[0022] It should be understood that terms such as “having,” “comprising,” and “including” as used herein do not exclude the presence or addition of one or more other elements or combinations thereof.

[0023] A water-mechanism collaborative intelligent management system suitable for multi-crop slopes includes an adjustable wheel track self-propelled sprinkler irrigation machine, a slope adaptive drip irrigation unit, and an intelligent sensing and control module. The adjustable wheel track self-propelled sprinkler irrigation machine includes a wheel track adjustment mechanism, a rubber track walking mechanism, a fertilizer storage tank, and a fault self-diagnosis module. The fault self-diagnosis module monitors the motor speed and pipeline pressure in real time. When the motor speed fluctuates by more than ±5% or the pipeline pressure fluctuates by more than ±0.2MPa, the module automatically stops the machine and pushes the fault code to the intelligent sensing and control module. The slope adaptive drip irrigation unit includes drip irrigation pipelines, slope sensors, and electromagnetic flow regulating valves. The interval of the drip irrigation pipelines is set according to the crop row spacing. The slope sensors collect the slope angle in real time, and the electromagnetic flow regulating valves adjust the water output according to the slope angle. The intelligent sensing and control module includes a soil moisture sensor, a weather station, a soil fertility sensor, and an embedded controller. Soil moisture sensors are buried at a depth of 20cm to 30cm, and soil fertility sensors are buried at a depth of 15cm to 20cm. The embedded controller uses wireless network technology to receive data in real time from soil moisture sensors, weather stations, and soil fertility sensors. The embedded controller presets irrigation thresholds for multiple crops and growth stages. It integrates slope sensor data, crop row spacing, and soil moisture information to make intelligent decisions and start adjustable wheel track self-propelled sprinkler irrigation machines or slope adaptive drip irrigation units to perform precise irrigation operations. Based on weather station data and soil fertility sensor data, the embedded controller generates fertilization ratio schemes and controls the fertilizer storage tank to perform variable fertilization. The embedded controller supports remote control of irrigation and fertilization amounts via mobile terminal applications.

[0024] Currently known slope irrigation systems mostly use fixed-wheel-track walking equipment with a relatively rigid chassis structure, making it difficult to adapt to uneven terrain and posing a risk of tipping over. Furthermore, they lack flexibility in handling different crop row spacings, often requiring manual intervention and resulting in low operational efficiency. Simultaneously, their irrigation control units often operate in isolation, relying solely on soil moisture as a single parameter to trigger irrigation, failing to consider multiple factors such as slope and crop type. This leads to uneven water distribution across different parts of the slope, with crops at higher elevations susceptible to drought while lower elevations are prone to waterlogging or runoff, resulting in low water and fertilizer utilization efficiency.

[0025] To overcome the aforementioned limitations, this system provides an integrated solution. The core of the system comprises three parts: an adjustable wheelbase self-propelled sprinkler irrigation machine, a slope-adaptive drip irrigation unit, and an intelligent sensing and control module. The adjustable wheelbase self-propelled sprinkler irrigation machine utilizes rubber tracks as its walking mechanism, significantly enhancing traction and maneuverability on muddy or soft slopes. The wheelbase adjustment mechanism is a prior art product, such as the high-clearance self-propelled chassis with adjustable wheelbase disclosed in publication number CN203126522U. The wheelbase adjustment range of this invention is 1.2m to 2.5m, with an adjustment accuracy of ±5mm. The adjustment motor is controlled by pulse signals sent by an embedded controller, achieving adaptive adjustment of crop row spacing. The wheelbase adjustment mechanism can be electrically adjusted by the operator via control terminal commands, allowing the wheelbase to match the actual row spacing requirements of different crops, effectively avoiding crop crushing. The sprinkler irrigation machine is equipped with an advanced fault self-diagnosis module, which monitors motor speed and irrigation pipeline pressure in real time to ensure the equipment operates within safe parameter ranges. When the motor speed fluctuates by more than ±5% or the pipeline pressure fluctuates by more than ±0.2%, the system will automatically shut down and send a specific fault code to the central control unit to prevent the equipment from continuing to operate under abnormal conditions, thereby avoiding potential damage.

[0026] The slope-adaptive drip irrigation unit consists of drip irrigation lines laid in the field, high-precision slope sensors, and electromagnetic flow control valves. The spacing of the drip irrigation lines is strictly pre-set according to the crop row spacing. The slope sensors collect real-time data on the slope's tilt angle, and the electromagnetic flow control valves adjust their opening and closing based on this real-time angle information to achieve precise control of irrigation water volume. For example, in areas with steeper slopes, the flow rate per unit time is automatically increased to compensate for the increased water flow velocity, thereby striving to achieve a uniform distribution of irrigation water across the entire slope. More specifically, when the slope sensor collects an angle of α°, the flow rate Q (L / h) of the electromagnetic flow control valve is calculated using the formula Q=Q0×(1+0.02α), where Q0 is the standard flow rate at a slope of 0° (preset to 50L / h). When α≥25°, the upper limit of the flow rate is set to 120L / h to prevent water accumulation in the lower slope area.

[0027] The embedded controller is equipped with various crops, including but not limited to corn, wheat, citrus, and strawberry. It presets corresponding row spacing adaptation ranges (corn 1.0m~1.2m, wheat 0.15m~0.2m, citrus 2.5m~3.0m, strawberry 0.3m~0.4m) and irrigation thresholds and nutrient requirement thresholds for each growth stage (seedling stage, growth stage, maturity stage) for different crops.

[0028] The intelligent sensing and control module, acting as the system's brain, is responsible for coordination and decision-making. This module integrates soil moisture and fertility sensors buried at different depths, a small weather station, and an embedded controller. The embedded controller continuously receives data from the sensors via a wireless network. It has pre-stored irrigation water requirement thresholds for various crops at different growth stages. The controller comprehensively analyzes real-time slope information, the currently planted crop type and row spacing data, and soil moisture content data to ultimately decide whether to activate the sprinkler irrigation system or the drip irrigation unit to perform the irrigation task. Simultaneously, it can generate scientific fertilizer formulas based on meteorological information and soil nutrient data, and control the fertilizer storage tank on the sprinkler irrigation system to perform variable-rate fertilization. Furthermore, users can remotely set or adjust irrigation and fertilizer application parameters via a mobile application, greatly improving management convenience.

[0029] Furthermore, the self-diagnosis module of the adjustable wheel track self-propelled sprinkler also integrates a GPS positioning module. The positioning accuracy of the GPS positioning module is within 10m. When the self-diagnosis module pushes a fault code, it simultaneously sends the real-time location information of the sprinkler. After receiving the fault code and location information, the embedded controller automatically generates a maintenance instruction containing the location of the faulty equipment and the fault type. The maintenance instruction is pushed to the mobile terminal of the maintenance personnel through a 4G or 5G network.

[0030] In the management of existing large-scale sloping plantations, intelligent irrigation systems, through integrated sensors and data analytics, can help maintenance personnel quickly locate the specific location of faulty equipment, thereby improving the maintenance efficiency of irrigation equipment. Traditional fault alarms typically only include the equipment number or a general area description, requiring maintenance personnel to rely on personal experience or familiarity with the plantation to locate the equipment, especially during periods of vigorous crop growth when equipment is obscured, making location even more difficult. This results in excessively long times from the onset of a fault to the arrival of maintenance personnel on-site, significantly impacting the continuity and efficiency of irrigation operations, and ultimately leading to unnecessary production losses.

[0031] To address the challenge of slow response times in locating and repairing faulty equipment, this implementation plan expands the functionality of the fault self-diagnosis module. This module integrates a GPS positioning chip with a positioning accuracy within 10 meters, transforming it into an intelligent terminal with spatial location awareness. Once the fault self-diagnosis module detects abnormal motor speed or unstable pipeline pressure and triggers the automatic shutdown mechanism, its response strategy has been upgraded from simply pushing fault codes to a more comprehensive operation. The module immediately invokes the integrated GPS positioning function to obtain the real-time latitude and longitude coordinates of the adjustable wheelbase self-propelled sprinkler, binds this location information with the fault code and equipment number, and sends it to the system's intelligent sensing and control module via the equipment's wireless communication unit.

[0032] Upon receiving this information, the embedded controller in the system automatically activates its built-in maintenance scheduling program. The program first parses the fault code to determine the fault type, then converts the received latitude and longitude coordinates into a specific geographical location description within the park. Subsequently, the controller automatically generates a structured maintenance instruction that not only clearly indicates the exact location of the faulty equipment but also describes the fault symptoms in detail. This complete maintenance instruction is then pushed directly as a message to the smartphone or dedicated mobile terminal of the on-duty maintenance personnel via the 4G or 5G mobile network covering the park.

[0033] Maintenance personnel can utilize the precise location navigation function of mobile terminals to quickly reach the site of faulty equipment, while simultaneously preparing the necessary tools and spare parts in advance based on the fault description. This efficient workflow significantly reduces troubleshooting and on-site repair preparation time, thereby improving maintenance efficiency and the reliability of the irrigation system.

[0034] Furthermore, the embedded controller is also connected to a local data storage module, which records the historical fault codes, maintenance records, and location history information of each adjustable wheel track self-propelled sprinkler. The maintenance instructions also include the three most recent fault records and the most recent maintenance record of the corresponding sprinkler. If the fault self-diagnosis module detects that the same fault code reappears, the embedded controller adds a warning signal to the maintenance instructions. The maintenance records include the name of the maintenance component, maintenance method, maintenance start and end time, and replacement part model.

[0035] In existing equipment maintenance procedures, when maintenance personnel arrive at the fault site, they typically only obtain the fault code currently reported by the equipment, knowing nothing about its past operating conditions and maintenance history. This lack of information leads to a lack of continuity in maintenance decisions, making it difficult for maintenance personnel to determine whether the current fault is an isolated event, a long-standing problem, or a legacy issue that was not fully resolved in the previous maintenance. For example, when faced with a recurring fault, the lack of historical records may cause maintenance personnel to perform only routine procedures without conducting in-depth investigations into the root cause, resulting in repeated fault occurrences and increased equipment downtime and maintenance costs.

[0036] To address the lack of historical information for maintenance personnel, this implementation plan connects a local data storage module to the embedded controller. This module continuously operates, meticulously recording historical fault codes, maintenance details, and location changes encountered by each adjustable wheelbase self-propelled sprinkler at every stage of its lifecycle, thereby constructing a comprehensive equipment health record.

[0037] When the fault self-diagnosis module detects an anomaly again and pushes a fault code and real-time location information, the embedded controller first initiates a data query and association process before generating a maintenance instruction. The controller accesses the local data storage module, using the faulty device's ID as an index, to quickly retrieve the device's three most recent fault records and the most recent detailed maintenance record. Subsequently, the system automatically integrates this valuable historical information into the newly generated maintenance instruction, making the issued instruction no longer an isolated current event report, but an important maintenance work order containing historical context.

[0038] The maintenance personnel receive a detailed instruction on their mobile terminal. They not only know the current location of the equipment and the nature of the malfunction, but also clearly understand what problems the equipment has encountered in the past and how it was previously repaired. More importantly, the system includes a simple logical check: if the controller detects that the currently received fault code is identical to a fault code in the equipment's historical records, it will automatically attach a prominent warning indicator when generating the maintenance instruction.

[0039] This warning sign is designed to remind maintenance personnel that the equipment may have an unresolved defect or that a component is nearing failure, requiring a more thorough and detailed inspection and handling of the equipment during this maintenance process. This avoids a one-sided approach to the fault and enhances the focus and completeness of the maintenance work.

[0040] Furthermore, the embedded controller also integrates a fault mode analysis module; the fault mode analysis module analyzes the correlation between fault codes based on historical fault codes and maintenance record data, using an association rule algorithm. When the fault mode analysis module receives a new fault code, it automatically matches other potentially related system components and generates a preventative checklist. The preventive inspection checklist is pushed to the maintenance personnel's mobile terminal along with the maintenance instructions; The preventative inspection checklist includes the names of the components to be inspected, the inspection methods, and any potential faults that may exist.

[0041] In complex agricultural irrigation equipment, faults between different system components often have potential correlations. However, current fault diagnosis technologies typically only address isolated reported fault codes. Upon arrival at the site, maintenance personnel rely primarily on their personal experience and expertise to determine whether further inspection of related components is necessary. This method is highly dependent on the technical skill level of the maintenance personnel. Inexperienced personnel are prone to overlooking potential correlations hidden behind the main fault, leading to incomplete repairs. This can result in different component failures caused by the same underlying reason recurring quickly, trapping the equipment in a vicious cycle of repeated repairs.

[0042] To overcome the limitations of single fault codes and achieve a shift from reactive maintenance to systematic preventative maintenance, this implementation plan integrates a fault mode analysis module within the embedded controller. The core function of this module is to deeply mine the component fault correlations hidden within accumulated historical maintenance data. Its operation does not rely on human experience but rather on an automated analysis of historical fault codes and maintenance record datasets based on association rule algorithms.

[0043] Specifically, this module runs continuously in the background. Once the fault self-diagnosis module sends a new fault code to the embedded controller, the fault mode analysis module is activated. It first takes the received current fault code as a key input, and then performs a rapid matching query in its effective association rule base, which it has learned through algorithms. This rule base reveals, for example, the potential pattern that component B often has a higher risk of failure when fault A occurs.

[0044] After the matching query is completed, if other components that are significantly related to the current fault code are found, the fault mode analysis module will automatically generate a preventative checklist. This checklist is no longer a simple fault description, but a proactive action guide. It details the names of related components that maintenance personnel are advised to check additionally, the specific inspection methods for each component, and the potential fault modes that these components may have.

[0045] Ultimately, this highly valuable preventative checklist will be combined with routine maintenance instructions and pushed to the maintenance personnel's mobile devices. Using this checklist, maintenance personnel can address known faults while simultaneously performing pre-emptive checks and maintenance on related components that may be about to fail, thus nipping potential problems in the bud and effectively preventing subsequent cascading downtime. This significantly improves the overall reliability and lifespan of the equipment.

[0046] Furthermore, the specific implementation of the association rule algorithm in the fault mode analysis module includes: The data preprocessing unit organizes historical fault codes into transaction datasets by device number and time sequence. Each transaction contains all fault codes recorded in a single maintenance event. The frequent itemset mining unit uses the FP-Growth algorithm to scan the transaction dataset and find frequent fault code itemsets with support between 2% and 5%. The rule generation unit generates association rules with a confidence level between 60% and 80% based on frequent itemsets; The rule filtering unit uses the lift index to evaluate the effectiveness of association rules, and only retains association rules with a lift greater than one for the generation of preventive checklists; The rule filtering unit automatically recalculates the support, confidence, and lift of all associated rules every month. When the confidence of a rule drops by more than 20% or the support drops by more than 50%, the rule is automatically removed from the valid rule base.

[0047] In existing technologies using association rule algorithms for equipment fault prediction, a common problem is the simplistic setting of algorithm parameters and the static, unchanging rule base. This often leads to the algorithm discovering a large number of seemingly relevant but actually meaningless rules, such as combinations of faults occurring only by chance, or rules that appear simultaneously but have no causal relationship. If maintenance personnel rely on these low-value rules for preventative checks, they will not only waste a lot of time but may also gradually ignore the checklist's prompts due to frequent false alarms, thus significantly reducing the effectiveness of the preventative maintenance mechanism.

[0048] To ensure the high reliability and practicality of the generated preventative checklist, this implementation plan features a refined process design and strict parameter control for the association rule algorithm in the fault mode analysis module. The algorithm begins with the data preprocessing unit, which cleans and organizes the massive amounts of historical maintenance data. A structured transaction dataset is constructed based on single maintenance events for each device, ensuring that each transaction record clearly reflects all fault codes processed during the maintenance process. The transaction dataset is stored in CSV format, with fields including device number, maintenance time, fault code 1, fault code 2…fault code n. The minimum support calculation basis for the FP-Growth algorithm is the total number of maintenance transactions in the past 12 months. During data scanning, frequent itemsets are cached in memory to reduce disk I / O operations and improve mining efficiency.

[0049] Subsequently, in the data mining process, the frequent itemset mining unit utilizes the FP-Growth algorithm to efficiently analyze the transaction dataset. The FP-Growth algorithm, by constructing an FP-tree, requires only two scans of the dataset, significantly improving data processing efficiency and reducing computational complexity through data compression. Unlike simply counting all combinations, this unit sets a strict support threshold range, identifying only fault code combinations with an occurrence frequency between 2% and 5% as valuable frequent itemsets, effectively filtering out invalid combinations that are too rare or too common. The support range of 2%-5% is based on statistics from equipment maintenance data over the past 12 months (the probability of multiple fault codes occurring concurrently in a single fault is concentrated in this range); the confidence range of 60%-80% is determined through fault association rule verification experiments.

[0050] After obtaining high-quality frequent itemsets, the rule generation unit begins its work. It generates candidate association rules based on these itemsets and calculates the confidence level of each rule. This unit sets another threshold, retaining only rules with a confidence level between 60% and 80%, ensuring that the rules have sufficient reliability without overfitting historical data.

[0051] Finally, the rule selection unit introduces a lift metric as the final evaluation standard to measure whether the correlation between fault codes in the rules is stronger than randomness. Only rules with a lift greater than one are proven to have practical correlation value and are included in the final effective rule base to support the generation of preventive checklists.

[0052] More importantly, the entire rule base is not static. The system automatically initiates a rule recalculation process every month to re-evaluate the support, confidence, and lift of all existing rules.

[0053] Once the system detects that the confidence level of a rule has dropped by more than 20% or the support level has dropped by more than 50%, indicating that the effectiveness of the rule has been significantly weakened or the failure mode it represents has changed, the system will automatically remove it from the valid rule base. This dynamic update mechanism ensures that preventative maintenance recommendations remain highly relevant to the current actual operating status of the equipment, continuously providing maintenance personnel with accurate and reliable action guidelines.

[0054] Furthermore, the embedded controller is provided with a device decision logic unit; The device decision logic unit receives slope sensor data, crop row spacing type data, and soil moisture content data; The equipment decision logic unit first determines the equipment type based on slope sensor data: When the slope sensor data is less than 10°, the decision is made to start the adjustable wheel track self-propelled sprinkler irrigation machine. When the slope sensor data is greater than or equal to 10° and less than or equal to 25°, the decision is made to activate the slope adaptive drip irrigation unit. Subsequently, the equipment decision logic unit makes irrigation intensity decisions based on soil moisture content data: When the soil moisture content is more than 5 percentage points lower than the threshold for the corresponding crop growth period, if the decision has been made to start the adjustable wheel track self-propelled sprinkler irrigation machine, then control it to perform 130% of the standard irrigation volume. If the decision has been made to activate the slope adaptive drip irrigation unit, then control it to operate for 140% of the standard irrigation duration; The equipment decision logic unit reassesses the soil moisture content data and updates the irrigation intensity control instructions every 30 minutes.

[0055] Existing control strategies for slope irrigation systems attempt to comprehensively consider multiple parameters, such as topographic conditions and crop water requirements, to achieve efficient water use and water-saving agriculture. However, this comprehensive judgment method is prone to rule conflicts, thus requiring scientific management measures and optimization techniques, such as referencing ecological water requirement studies at the oasis irrigation scale and irrigation and drainage design in the Loess Plateau hilly and gully region, to improve decision-making efficiency. For example, the system may simultaneously receive instructions indicating that the slope is suitable for sprinkler irrigation but the soil is severely water-deficient. If the processing logic is inappropriate, it may issue instructions to start both sprinkler and drip irrigation units simultaneously, physically causing interference between equipment or wasting water resources. This multi-parameter parallel decision-making mode lacks priority classification, making it difficult to guarantee the reliability of decisions and operational safety in complex environments.

[0056] As a further optimization of the aforementioned system with fault diagnosis and location functions, this implementation scheme adds a device decision logic unit to the embedded controller. The core design concept of this unit is to decouple the complex multi-parameter decision-making process into two clear serial steps. This unit continuously receives terrain angle data from the slope sensor, pre-set crop row spacing type data, and real-time moisture content data transmitted from the soil moisture sensor.

[0057] Its decision-making process is strictly based on slope, a core parameter related to operational safety. The unit has a pre-set, clear slope threshold: when the slope sensor data is less than 10°, the terrain is considered flat, and the adjustable-wheel-track self-propelled sprinkler system is activated for efficient operation; when the slope data is greater than or equal to 10° and less than or equal to 25°, the terrain is considered gentle to moderate, and the adaptive drip irrigation unit, more suitable for slopes, is activated, ensuring the fundamental safety of equipment selection.

[0058] After determining the appropriate equipment type, the decision-making logic proceeds to the second step: adjusting irrigation intensity based on the severity of drought. The equipment decision-making logic unit compares real-time soil moisture content with the standard water requirement threshold for the current crop growth stage. When the soil moisture content is detected to be more than five percentage points below the threshold, indicating moderate water shortage, the system activates an enhanced irrigation mode. If the first decision was to activate a sprinkler irrigation system, it is controlled to perform 130% of the standard irrigation volume; if the first decision was to activate a drip irrigation unit, it is controlled to perform 140% of the standard irrigation duration. This differentiated intensity control based on equipment characteristics aims to quickly alleviate drought conditions.

[0059] To ensure that irrigation strategies adapt to environmental changes, the device's decision-making logic unit does not make decisions all at once, but is set to automatically reassess every 30 minutes. It acquires the latest soil moisture data and adjusts irrigation intensity control commands in real time accordingly, achieving safe and dynamically precise irrigation management, effectively avoiding the risks of equipment conflicts and water resource misallocation. This cycle is set based on the response speed of the soil moisture sensor (≤5 minutes) and the soil moisture infiltration rate on slopes (approximately 2%–3% change in moisture every 30 minutes), ensuring timely capture of soil moisture changes. During the assessment, data from three evenly distributed soil moisture sensors in the area are collected, and the average value is used as the judgment basis to avoid the influence of single sensor errors on the decision.

[0060] Furthermore, the device decision logic unit is also connected to a data reliability assessment module; The data reliability assessment module monitors the data output status of the slope sensor and soil moisture sensor in real time; When the standard deviation of three consecutive slope sensor readings is greater than 1°, the sensor data is deemed unreliable. When the coefficient of variation of soil moisture sensor data collected over 10 consecutive minutes is greater than 15%, the sensor data is deemed unreliable. When any sensor data is determined to be unreliable, the data reliability assessment module sends a data failure signal to the device decision logic unit. Once the equipment's decision logic unit receives a data failure signal, it will automatically switch to a preset irrigation mode based on crop growth period and weather station data.

[0061] In the preset irrigation mode, the equipment decision logic unit performs irrigation operations according to the standard irrigation amount for the current crop growth period, and reassesses the reliability status of sensor data every 6 hours.

[0062] In existing smart irrigation systems, control units typically rely directly on raw data from sensors for decision-making, lacking an effective mechanism to verify the reliability of the data itself. When sensors malfunction, experience external interference, or become loosely installed, resulting in severely distorted measurements—for example, a slope sensor output fluctuates drastically or a soil moisture sensor reading remains stagnant for an extended period—the system may make irrigation decisions completely deviating from reality based on this erroneous data. This could lead to risks such as incorrectly starting sprinkler systems on steep slopes or continuing to irrigate at normal rates even when drought conditions are severe, directly impacting crop growth and making it difficult to assign blame.

[0063] As a further improvement to the aforementioned system with intelligent decision-making and fault management functions, this implementation scheme connects a dedicated data reliability assessment module to the device decision logic unit within the embedded controller. The core responsibility of this module is to act as a "sentinel" for sensor data quality, performing real-time online diagnostics on the raw data from key sensors it receives, rather than blindly accepting it.

[0064] This module has clearly defined data quality criteria. For slope sensors, the module continuously calculates the standard deviation of the three most recent data acquisitions. If the standard deviation exceeds 1°, it indicates abnormal fluctuations in the slope readings and unacceptable data stability, leading the module to determine that the sensor's current data is unreliable. For soil moisture sensors, the module calculates the coefficient of variation of the data acquired over ten consecutive minutes. If this coefficient exceeds 15%, it indicates excessive variation in moisture readings, also resulting in unreliable data.

[0065] Once any sensor is determined to have failed, the data reliability assessment module immediately sends a high-priority digital signal—a data failure signal—to the equipment decision-making logic unit. Upon receiving this signal, the equipment decision-making logic unit does not enter a complex multi-parameter decision-making process, but automatically executes a pre-set, conservative yet reliable contingency plan. It immediately switches to a preset irrigation mode based on crop growth stage and weather station data.

[0066] In this preset mode, the system temporarily ignores readings from faulty sensors and instead executes timed and quantitative irrigation according to the standard irrigation formula for the current crop growth stage. This avoids extreme situations such as insufficient or excessive irrigation due to erroneous data. Simultaneously, the system does not abandon tracking the faulty sensor. This module automatically reassesses the data output status of the faulty sensor every six hours. Once the data quality stabilizes and passes the reliability check again, the system automatically deactivates the preset mode and restores its intelligent decision-making function, achieving a balance between safety and intelligence.

[0067] Furthermore, the embedded controller is equipped with a fertilization decision unit; The fertilization decision-making unit performs precise fertilization based on the 24-hour precipitation probability data provided by the meteorological station and the nitrogen, phosphorus, and potassium content data provided by the verified soil fertility testing instrument.

[0068] The fertilization decision unit has built-in threshold ranges for nutrient requirements at different growth stages of different crops; When the content of a certain element in the soil is lower than the lower limit of the threshold required by the corresponding crop at the current growth stage, the fertilization decision unit generates a supplementary formula for that element. If the probability of precipitation exceeds 50% within the next 24 hours, the fertilization decision-making unit will postpone the fertilization plan until the precipitation event has completely ended, based on a scientific meteorological assessment, in order to avoid fertilizer loss due to rainwater erosion and ensure the effectiveness of fertilization.

[0069] The fertilization decision unit generates fertilization control instructions and sends them to the adjustable wheel track self-propelled sprinkler irrigation machine to control the mixing ratio of nitrogen, phosphorus and potassium fertilizers in its fertilizer storage tank, with the mixing accuracy controlled within ±3%. The fertilization decision unit dynamically adjusts the fertilization interval to 3–7 days based on the rate of change of soil fertility sensor data.

[0070] The rate of change of the soil fertility sensor data is the difference in the daily variation of soil nitrogen, phosphorus, and potassium content within a unit time (24h). When the absolute value of the daily variation of a certain element content is less than 0.3 mg / kg, the fertilization interval is adjusted to 7 days; when the absolute value of the variation is between 0.3 mg / kg and 0.8 mg / kg, the fertilization interval is adjusted to 5 days; and when the absolute value of the variation is greater than 0.8 mg / kg, the fertilization interval is adjusted to 3 days.

[0071] In existing slope fertilization management, decisions are typically made primarily based on real-time soil nutrient content testing, lacking comprehensive consideration of crop stage-specific nutrient requirements and even less incorporating the influence of meteorological factors. This static fertilization strategy has significant drawbacks. For example, even if a deficiency of a certain element in the soil is detected and replenished immediately, if heavy rainfall occurs shortly after fertilization, the newly applied fertilizer is easily washed away or leached, leading to a significant reduction in fertilizer utilization. This results in economic waste and may also pose a risk of non-point source pollution. Simply relying on soil testing for fertilization is insufficient to address the complex and ever-changing external conditions in slope environments.

[0072] As a functional extension of the aforementioned system integrating irrigation control, fault diagnosis, and data reliability assessment, this implementation scheme incorporates a dedicated fertilization decision-making unit within the embedded controller. This unit is designed to upgrade fertilization decisions from static to dynamic, and from single-factor to multi-factor fusion. It continuously receives two key external data streams: one is the 24-hour precipitation probability forecast data provided by the weather station; the other is the real-time nitrogen, phosphorus, and potassium element content data transmitted from soil fertility sensors.

[0073] This unit pre-stores the threshold ranges of nitrogen, phosphorus, and potassium nutrient requirements for different crops at various major growth stages. Its decision-making logic first compares the real-time monitored soil nutrient data with the current crop growth stage's requirement thresholds. When it detects that the content of a certain element in the soil is below the lower limit of the corresponding requirement threshold, the fertilization decision unit generates a supplementary formula for that element.

[0074] However, the decision-making process doesn't end there. The unit immediately retrieves the precipitation probability data for the next 24 hours for a second assessment. This is a crucial risk control step: when the probability of precipitation in the next 24 hours is determined to be greater than 50%, the unit does not cancel fertilization, but intelligently postpones the execution of the fertilization plan until after the predicted precipitation event has completely ended, thus effectively avoiding the risk of fertilizer being washed away immediately after application.

[0075] Finally, the fertilization decision unit generates a detailed fertilization control instruction and sends it to the adjustable wheelbase self-propelled sprinkler irrigation machine. This instruction ensures the optimal ratio of nitrogen, phosphorus, and potassium nutrients required for crop growth by precisely controlling the mixing ratio of each mother liquor fertilizer in the fertilizer storage tank, while maintaining the mixing accuracy within ±3% to meet the optimal growth needs of the crop. Furthermore, this unit does not have a fixed fertilization cycle; it dynamically judges the rate of soil nutrient consumption based on the data change rate fed back by soil fertility sensors, thus intelligently adjusting the fertilization interval within the range of three to seven days, achieving on-demand fertilization based on the actual consumption of the crop.

[0076] Furthermore, the fertilization decision unit is also connected to a fertilization effect feedback and evaluation module; The fertilization effect feedback and evaluation module is activated after each fertilization plan is executed, and continuously monitors the changes in soil fertility sensor data. This module sets the expected rate of increase in nitrogen concentration to a threshold of 0.5 mg / kg to 1.0 mg / kg per hour; When the actual monitored rate of increase in nitrogen concentration is lower than the lower limit of the expected threshold, the fertilization effect feedback evaluation module sends a signal of insufficient fertilizer effect to the fertilization decision unit. After receiving a signal of insufficient fertilizer effectiveness, the fertilization decision-making unit automatically initiates a secondary precision topdressing procedure. The second precision topdressing program is then initiated, adding 20% ​​to 30% of the originally planned fertilizer amount in liquid form through a self-propelled sprinkler irrigation system with adjustable wheel spacing. The fertilization effect feedback and evaluation module continues to monitor for 2 hours after additional fertilization. If the fertilizer effect is still insufficient, it will generate a sensor calibration prompt.

[0077] In existing sloping land fertilization management, decisions are typically made primarily based on real-time soil nutrient content testing, lacking comprehensive consideration of crop stage-specific nutrient requirements and even less incorporating the influence of meteorological environmental factors. This static fertilization strategy has significant drawbacks. For example, even if element deficiencies in the soil can be quickly detected and replenished in a timely manner, if heavy rainfall occurs shortly after fertilization, the newly applied fertilizer may still be washed away or leached by rainwater, significantly reducing fertilizer utilization. This not only causes economic losses but may also trigger non-point source pollution risks. Simply relying on "soil testing for fertilization" is insufficient to cope with the complex and ever-changing external conditions in sloping land environments.

[0078] As a functional extension of the aforementioned system integrating irrigation control, fault diagnosis, and data reliability assessment, this implementation scheme incorporates a dedicated fertilization decision-making unit within the embedded controller. This unit is designed to upgrade fertilization decisions from static to dynamic, and from single-factor to multi-factor fusion. It continuously receives two key external data streams: one is the 24-hour precipitation probability forecast data provided by the weather station; the other is the real-time nitrogen, phosphorus, and potassium element content data transmitted from soil fertility sensors.

[0079] This unit pre-stores the threshold ranges of nitrogen, phosphorus, and potassium nutrient requirements for different crops at various major growth stages. Its decision-making logic first compares real-time monitored soil nutrient data with the threshold requirements for the current crop at a specific growth stage, taking into account the differences in nutrient requirements at different growth and development stages. When the content of a certain element in the soil is found to be below the lower limit of its corresponding threshold, the fertilization decision unit generates a supplementary formula for that element.

[0080] However, the decision-making process doesn't end there. The unit immediately retrieves the precipitation probability data for the next 24 hours for a second assessment. This is a crucial risk control step: when the probability of precipitation in the next 24 hours is determined to be greater than 50%, the unit does not cancel fertilization, but intelligently postpones the execution of the fertilization plan until after the predicted precipitation event has completely ended, thus effectively avoiding the risk of fertilizer being washed away immediately after application.

[0081] Finally, the fertilization decision unit generates a detailed fertilization control instruction and sends it to the adjustable-wheel-track self-propelled sprinkler irrigation machine. This instruction precisely controls the mixing ratio of each mother liquor fertilizer in its fertilizer storage tank to achieve the required nitrogen, phosphorus, and potassium ratio, with mixing accuracy controlled within ±3%. Furthermore, this unit does not have a fixed fertilization cycle; it dynamically judges the soil's nutrient consumption rate based on the rate of change in data from soil fertility sensors, intelligently adjusting the fertilization interval to within the range of three to seven days, achieving on-demand fertilization based on actual crop consumption.

[0082] Effect test To scientifically verify the comprehensive advantages of the system of the present invention (hereinafter referred to as the "intelligent collaborative system") compared with traditional irrigation methods in multi-crop environments on slopes, the following experiments were conducted.

[0083] I. Experimental Design 1. Experimental site and crop arrangement Wheat experimental area: slope 5°, row spacing 0.2m; Citrus experimental area: slope 20°, row spacing 2.5m.

[0084] 2. Experimental group and control group Experimental group: Using the "Moisture-Mechanical Collaborative Intelligent Management System" described in this invention; Control group: Irrigation and fertilization were decided by traditional fixed wheel gauge sprinkler irrigation machine and human experience.

[0085] 3. Measurement Indicators Water use efficiency = crop yield / total irrigation water consumption; Fertilizer utilization rate = amount of nutrients absorbed by the crop / total amount of fertilizer applied; Mean time to resolve equipment failures; Crop yield per mu (kg / mu).

[0086] II. Experimental Data Results Table 1 Measurement Index Results Experimental results show that the system of this invention exhibits significant comprehensive advantages compared to traditional irrigation methods. The intelligent system, through slope-adaptive flow regulation and soil moisture feedback mechanisms, effectively avoids waterlogging at the foot of the slope and drought at the top, increasing water use efficiency for wheat and citrus cultivation by 50% and 87.5%, respectively. Simultaneously, the system dynamically adjusts the timing and ratio of fertilization based on soil fertility sensors and meteorological data, avoiding fertilizer waste caused by rainfall erosion and over-application, thus increasing fertilizer utilization efficiency by 17%. Furthermore, the system's integrated fault self-diagnosis, GPS positioning, and maintenance record push functions significantly reduce the average fault handling time to 20%–25% of traditional methods, greatly improving operational efficiency. Ultimately, through the combined effects of precise water and fertilizer management, improved irrigation uniformity, and enhanced equipment reliability, the yield per mu (a Chinese unit of area, approximately 0.067 hectares) of wheat and citrus significantly increased by 21.4% and 26.1%, respectively, fully demonstrating the system's superior performance in improving the efficiency of agricultural resource utilization and increasing yields on slopes.

[0087] Although the embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details.

Claims

1. A water-mechanism collaborative intelligent management system suitable for multi-crop slopes, comprising an adjustable wheel track self-propelled sprinkler irrigation machine, a slope-adaptive drip irrigation unit, and an intelligent sensing and control module, characterized in that: The adjustable wheel track self-propelled sprinkler irrigation machine includes a wheel track adjustment mechanism, a rubber track walking mechanism, a fertilizer storage tank, and a fault self-diagnosis module. The fault self-diagnosis module monitors the motor speed and pipeline pressure in real time. When the motor speed fluctuates by more than ±5% or the pipeline pressure fluctuates by more than ±0.2MPa, the module automatically stops the machine and pushes the fault code to the intelligent sensing and control module. The slope adaptive drip irrigation unit includes drip irrigation pipelines, slope sensors, and electromagnetic flow regulating valves. The interval of the drip irrigation pipelines is set according to the crop row spacing. The slope sensors collect the slope angle in real time, and the electromagnetic flow regulating valves adjust the water output according to the slope angle. The intelligent sensing and control module includes a soil moisture sensor, a weather station, a soil fertility sensor, and an embedded controller. Soil moisture sensors are buried at a depth of 20cm to 30cm, and soil fertility sensors are buried at a depth of 15cm to 20cm. The embedded controller receives data from soil moisture sensors, weather stations, and soil fertility sensors via a wireless network; The embedded controller incorporates irrigation thresholds for various crops and different growth stages. It can intelligently decide and activate adjustable wheel track self-propelled sprinkler irrigation machines or slope-adaptive drip irrigation units by integrating slope sensor data, crop row spacing type, and soil moisture content information to carry out precise irrigation operations. The embedded controller generates fertilization ratio schemes based on weather station data and soil fertility sensor data and controls the fertilizer storage tank to perform variable fertilization. The embedded controller also supports remote control of irrigation and fertilization amounts via mobile terminal applications.

2. The water-mechanical collaborative intelligent management system for multi-crop slopes according to claim 1, characterized in that, The self-diagnosis module of the adjustable wheel track self-propelled sprinkler also integrates a GPS positioning module. The positioning accuracy of the GPS positioning module is within 10m. When the self-diagnosis module pushes a fault code, it simultaneously sends the real-time location information of the sprinkler. After receiving the fault code and location information, the embedded controller automatically generates a maintenance instruction containing the location of the faulty equipment and the fault type. The maintenance instruction is pushed to the mobile terminal of the maintenance personnel through a 4G or 5G network.

3. The water-mechanical collaborative intelligent management system for multi-crop slopes according to claim 2, characterized in that: The embedded controller is also connected to a local data storage module, which records the historical fault codes, maintenance records, and location history information of each adjustable wheel track self-propelled sprinkler. The maintenance instructions also include the three most recent fault records and the most recent maintenance record of the corresponding sprinkler. When the fault self-diagnosis module pushes the same fault code repeatedly, the embedded controller adds a warning indicator to the maintenance instructions. The maintenance records include the name of the maintenance component, maintenance method, maintenance start and end time, and replacement part model.

4. The water-mechanical collaborative intelligent management system for multi-crop slopes according to claim 3, characterized in that, The embedded controller also integrates a fault mode analysis module; the fault mode analysis module analyzes the correlation between fault codes based on historical fault codes and maintenance record data, using an association rule algorithm. When the fault mode analysis module receives a new fault code, it automatically matches other potentially related system components and generates a preventative checklist. The preventive inspection checklist is pushed to the maintenance personnel's mobile terminal along with the maintenance instructions; The preventative inspection checklist includes the names of the components to be inspected, the inspection methods, and any potential faults that may exist.

5. The water-mechanical collaborative intelligent management system for multi-crop slopes according to claim 4, characterized in that, The specific implementation of the association rule algorithm in the fault mode analysis module includes: The data preprocessing unit organizes historical fault codes into transaction datasets by device number and time sequence. Each transaction contains all fault codes recorded in a single maintenance event. The frequent itemset mining unit uses the FP-Growth algorithm to scan the transaction dataset and find frequent fault code itemsets with support between 2% and 5%. The rule generation unit generates association rules with a confidence level between 60% and 80% based on frequent itemsets; The rule filtering unit uses the lift index to evaluate the effectiveness of association rules, and only retains association rules with a lift greater than one for the generation of preventive checklists; The rule filtering unit automatically recalculates the support, confidence, and lift of all associated rules every month. When the confidence of a rule drops by more than 20% or the support drops by more than 50%, the rule is automatically removed from the valid rule base.

6. The water-mechanical collaborative intelligent management system for multi-crop slopes according to claim 1, characterized in that, The embedded controller is equipped with a device decision logic unit; The device decision logic unit receives slope sensor data, crop row spacing type data, and soil moisture content data; The equipment decision logic unit first determines the equipment type based on slope sensor data: When the slope sensor data is less than 10°, the decision is made to start the adjustable wheel track self-propelled sprinkler irrigation machine. When the slope sensor data is greater than or equal to 10° and less than or equal to 25°, the decision is made to activate the slope adaptive drip irrigation unit. Subsequently, the equipment decision logic unit makes irrigation intensity decisions based on soil moisture content data: When the soil moisture content is more than 5 percentage points lower than the threshold for the corresponding crop growth period, if the decision has been made to start the adjustable wheel track self-propelled sprinkler irrigation machine, then control it to perform 130% of the standard irrigation volume. If the decision has been made to activate the slope adaptive drip irrigation unit, then control it to operate for 140% of the standard irrigation duration; The equipment decision logic unit reassesses the soil moisture content data and updates the irrigation intensity control instructions every 30 minutes.

7. The water-mechanical collaborative intelligent management system for multi-crop slopes according to claim 6, characterized in that, The device decision logic unit is also connected to a data credibility assessment module; The data reliability assessment module monitors the data output status of the slope sensor and soil moisture sensor in real time; When the standard deviation of three consecutive slope sensor readings is greater than 1°, the sensor data is deemed unreliable. When the coefficient of variation of soil moisture sensor data collected over 10 consecutive minutes is greater than 15%, the sensor data is deemed unreliable. When any sensor data is determined to be unreliable, the data reliability assessment module sends a data failure signal to the device decision logic unit. After receiving a data failure signal, the equipment decision logic unit automatically switches to a preset irrigation mode based on crop growth period and weather station data. In the preset irrigation mode, the equipment decision logic unit performs irrigation operations according to the standard irrigation amount for the current crop growth period, and reassesses the reliability status of sensor data every 6 hours.

8. The water-mechanical collaborative intelligent management system for multi-crop slopes according to claim 7, characterized in that, The embedded controller is equipped with a fertilization decision unit; The fertilization decision unit receives precipitation probability data for the next 24 hours from the meteorological station and nitrogen, phosphorus, and potassium content data from the soil fertility sensor. The fertilization decision unit has built-in threshold ranges for nutrient requirements at different growth stages of different crops; When the content of a certain element in the soil is lower than the lower limit of the threshold required by the corresponding crop at the current growth stage, the fertilization decision unit generates a supplementary formula for that element. When the probability of precipitation in the next 24 hours is greater than 50%, the fertilization decision-making unit will postpone the implementation of the fertilization plan until after the precipitation event ends; The fertilization decision unit generates fertilization control instructions and sends them to the adjustable wheel track self-propelled sprinkler irrigation machine to control the mixing ratio of nitrogen, phosphorus and potassium fertilizers in its fertilizer storage tank, with the mixing accuracy controlled within ±3%. The fertilization decision unit dynamically adjusts the fertilization interval to 3–7 days based on the rate of change of soil fertility sensor data.

9. The water-mechanical collaborative intelligent management system for multi-crop slopes according to claim 8, characterized in that, The fertilization decision-making unit is also connected to a fertilization effect feedback and evaluation module; The fertilization effect feedback and evaluation module is activated after each fertilization plan is executed, and continuously monitors the changes in soil fertility sensor data. This module sets the expected rate of increase in nitrogen concentration to a threshold of 0.5-1.0 mg / kg per hour; When the actual monitored rate of increase in nitrogen concentration is lower than the lower limit of the expected threshold, the fertilization effect feedback evaluation module sends a signal of insufficient fertilizer effect to the fertilization decision unit. After receiving a signal of insufficient fertilizer effectiveness, the fertilization decision-making unit automatically initiates a secondary precision topdressing procedure. The secondary precision topdressing process adds 20% to 30% of the original fertilizer amount in the form of liquid fertilizer through an adjustable wheel track self-propelled sprinkler irrigation machine. The fertilization effect feedback and evaluation module continues to monitor for 2 hours after additional fertilization. If the fertilizer effect is still insufficient, it will generate a sensor calibration prompt.

Images