Variable application system based on iko-pid parameter tuning

Abstract

The application discloses a variable pesticide application system based on IKO-PID parameter setting, and relates to the technical field of intelligent control of agricultural plant protection machinery, comprising a core controller, an execution module, a signal acquisition module and a power module. The core controller runs a PID control algorithm and automatically sets PID parameters through an improved K-means clustering optimization algorithm. The improvement includes population initialization strategy, population diversity enhancement strategy, local optimal escape strategy and convergence acceleration strategy. The execution module includes a rotating cage nozzle driving unit and a brushless water pump driving unit for atomization and delivery of pesticide liquid. The signal acquisition module acquires system state parameters in real time and feeds back, realizing closed-loop control. The application can automatically optimize and obtain global optimal PID parameters, significantly improve the control accuracy and response speed of the pesticide application system, effectively solve the problems of heavy spraying and missed spraying, improve the utilization rate of pesticides, and is suitable for variable pesticide application scenes such as fields, orchards and greenhouses.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for agricultural plant protection machinery, and in particular to a variable application system based on IKO-PID parameter tuning. Background Technology

[0002] Variable-rate pesticide application systems are a core component of precision agriculture technology. They adjust pesticide flow rates in real time to adapt to differences in crop growth and environmental changes, aiming to improve pesticide utilization and reduce chemical residues and environmental pollution. In existing technologies, PID controllers are widely used in the flow control stage of variable-rate pesticide application systems due to their simple structure and high reliability. However, the parameter tuning of traditional PID controllers often relies on manual experience or classical tuning methods (such as the Ziegler-Nichols method), which suffers from problems such as cumbersome tuning processes, poor adaptability, and susceptibility to system nonlinearity and time-varying characteristics. This leads to insufficient control accuracy and easily results in phenomena such as over-spraying and under-spraying during dynamic operations.

[0003] To address the aforementioned problems, some improvements have been proposed in the existing technology. For example, Chinese patent document CN112154996A discloses a dual-PID control precision variable dosing system and its feedback adjustment method. This system employs a dual-PID control structure, with the main PID controller regulating flow and the secondary PID controller performing pressure or motor speed compensation to improve system response speed and control stability. This solution improves upon the shortcomings of single-PID control to some extent, but its PID parameter tuning still relies on traditional methods, failing to fundamentally solve the global search and local convergence problems in parameter optimization. Specifically, traditional tuning methods struggle to automatically and efficiently obtain globally optimal PID parameters under complex and changing operating conditions, leading to decreased control performance when dealing with load disturbances and environmental changes, and still exhibiting defects such as large overshoot and high steady-state error.

[0004] In addition, most existing variable-rate application systems lack intelligent parameter self-tuning mechanisms, making it impossible to dynamically adjust control parameters based on real-time system status, which restricts their further application in precision agriculture. Summary of the Invention

[0005] This invention discloses a variable-rate application system based on IKO-PID parameter tuning, aiming to solve the technical problems in existing variable-rate application systems. Traditional parameter tuning methods rely on manual experience, making it difficult to adapt to complex working conditions. Existing optimization methods are prone to getting trapped in local optima, leading to over-spraying and under-spraying in dynamic operations. Furthermore, the lack of an intelligent self-tuning mechanism prevents the system from adjusting parameters in real time according to the system status, thus restricting the improvement of control accuracy and adaptability.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] The variable dosing system based on IKO-PID parameter tuning includes: a core controller, an execution module, a signal acquisition module, and a power supply module.

[0008] The core controller is used to run the control program and PID control algorithm, and output control signals to the execution module.

[0009] The execution module includes a rotary nozzle drive unit and a brushless water pump drive unit, used to perform atomization and delivery of the liquid medicine.

[0010] The signal acquisition module is used to collect system status parameters in real time and feed them back to the core controller to form closed-loop control.

[0011] The power module supplies power to all modules of the system.

[0012] The parameters of the PID control algorithm are tuned by an improved K-means clustering optimization algorithm. The improvements include a population initialization strategy, a population diversity enhancement strategy, a local optimum escape strategy, and a convergence acceleration strategy.

[0013] By introducing a PID controller tuned based on the improved K-means clustering optimization algorithm (IKO), a qualitative leap in control performance has been achieved. This intelligent algorithm integrates four strategies: optimal point set initialization, differential evolution, lens imaging reverse learning, and dynamic position guidance. It can automatically find the globally optimal PID parameters, achieve fast and accurate flow tracking and stable output, and fundamentally solve the problem of overspraying and underspraying in traditional pesticide application systems, thus significantly improving pesticide utilization.

[0014] In a preferred embodiment, the population initialization strategy is to generate an initial population using the optimal point set theory, so that the population is uniformly distributed within the parameter search space.

[0015] By initializing the population using the optimal point set theory, the algorithm ensures a uniform distribution of starting points within the parameter search space, avoids the search blind spots that may be caused by traditional random initialization, improves the algorithm's global exploration capability, reduces the risk of premature convergence, and lays a solid foundation for finding the globally optimal PID parameters.

[0016] In a preferred embodiment, the population diversity enhancement strategy is to introduce differential evolution operations during the algorithm iteration process.

[0017] The introduction of differential evolution effectively enhances population diversity, prevents the algorithm from getting stuck in local optima during iteration, improves the robustness and optimization efficiency of the algorithm, and ensures that the PID parameter tuning process can continuously explore better solutions and adapt to dynamically changing environmental conditions.

[0018] In a preferred scheme, the local optimum escape strategy is to use a lens imaging back-learning method to process the current optimal solution in order to explore potential better solution space regions.

[0019] The lens imaging back-learning method simulates the principle of optical reflection to learn from the current optimal solution, helping the algorithm escape the local optimum trap, expanding the search range of the solution space, and increasing the probability of finding a better combination of PID parameters, thereby improving the system control accuracy and stability.

[0020] In a preferred embodiment, the convergence acceleration strategy is to adopt a dynamic guidance method based on the global historical optimal solution, so that the search agent moves towards the optimal solution during the iteration process.

[0021] By utilizing global historical optimal information through a dynamic guidance method, the algorithm's convergence process is accelerated. The search agent can move quickly in a directional manner, reducing unnecessary iterations, improving parameter tuning efficiency, and enabling the system to reach a stable state more quickly. This approach is suitable for application scenarios with high real-time requirements.

[0022] In a preferred embodiment, the core controller controls the brushless motor in the rotary nozzle drive unit and the brushless water pump in the brushless water pump drive unit by outputting PWM signals.

[0023] Precise pulse width modulation via PWM signals enables fine control of the brushless motor and brushless water pump, ensuring accuracy and response speed in the atomization and delivery of the pesticide solution, avoiding flow fluctuations, and improving the uniformity of pesticide application and system reliability.

[0024] In a preferred embodiment, the signal acquisition module includes a flow sensor and a positioning module; the system also includes a wireless transmission module for data interaction and remote control with a remote cloud platform.

[0025] The system uses a flow sensor to monitor the flow rate of pesticide solution in real time, a positioning module to obtain the location information of the application, and a wireless transmission module to enable remote data interaction and remote control. This enhances the intelligence and traceability of the system, supports precision agricultural management, and facilitates real-time adjustment of application strategies and fault diagnosis.

[0026] The present invention also provides a method for tuning PID parameters in a variable dosage system, comprising the following steps:

[0027] S1: Based on the IKO-PID parameter tuning, a dynamic mathematical model of the variable-rate dosing system or its actuator is established, laying the theoretical foundation and providing an accurate simulation environment for the entire optimization process. By establishing the system's dynamic model, the real-world operating characteristics of the system can be accurately simulated in a computer without relying on expensive and time-consuming physical experiments. This avoids the equipment risks that may arise from direct debugging on physical machines and provides a safe, controllable, and repeatable testing platform for subsequent algorithm optimization.

[0028] S2: In a simulation environment, the variable-rate dosing system based on IKO-PID parameter tuning uses the performance index of the control system of the model as the fitness function. The improved K-means clustering optimization algorithm is used to automatically optimize the proportional, integral, and derivative parameters of the PID controller, quantifying the performance requirements of the control system as a fitness function, making the optimization objective very clear. By using the IKO algorithm, which integrates multiple improved strategies, for automatic optimization, it is possible to:

[0029] Overcoming the limitations of traditional tuning methods: It eliminates the over-reliance on engineers' experience and the blindness and suboptimal nature of manual trial and error.

[0030] Achieving global optimization: Intelligent algorithms can perform efficient searches within a large parameter space, effectively avoiding getting trapped in local optima, thereby finding a set of PID parameter combinations with globally optimal overall performance.

[0031] Improving tuning efficiency and accuracy: The automated optimization process greatly shortens the parameter tuning time and can obtain control parameters with accuracy far exceeding that of manual debugging, providing a fundamental guarantee for the system to achieve "fast, accurate and stable" control.

[0032] S3: After the algorithm of the variable drug delivery system based on IKO-PID parameter tuning has completed its iterative optimization, it outputs the optimal combination of PID parameters and writes it into the control program of the core controller, thus completing the closed loop from virtual simulation to actual application. The optimal solution obtained by the algorithm in the simulation environment is seamlessly connected to the actual hardware system, so that the theoretical optimization results are directly transformed into the improvement of actual control performance. This ensures that each variable drug delivery system carries deeply optimized and "tailor-made" control parameters, thereby fundamentally guaranteeing the consistency and superiority of the product's control performance.

[0033] In a preferred embodiment, the automatic optimization process sequentially executes the following steps: population initialization based on the set of best points, population update incorporating a differential evolution strategy, global exploration or local development selection based on location density probability, dynamic location guidance based on the global optimal solution, and optimal solution update based on lens imaging back learning.

[0034] Through a systematic optimization process, the integrity and efficiency of the PID parameter tuning process are ensured. Each step works collaboratively, balancing global exploration and local development, dynamically adjusting the search direction, and ultimately outputting the optimal PID parameter combination. This maximizes system control performance, enabling rapid flow tracking and stable output, fundamentally solving the problems of over-spraying and under-spraying.

[0035] As can be seen from the above, the variable dosing system based on IKO-PID parameter tuning provided by this invention has the following technical effects.

[0036] Firstly, by employing an improved K-means clustering optimization algorithm to automatically tune the PID controller parameters, global optimization of the control parameters is achieved, significantly improving the system's control accuracy and response speed, and fundamentally solving the problems of over-spraying and under-spraying caused by improper parameter tuning in traditional pesticide application systems.

[0037] Secondly, the improved algorithm integrates four major strategies: optimal point set initialization, differential evolution, lens imaging reverse learning, and dynamic position guidance. This effectively overcomes the shortcomings of traditional optimization methods that are prone to getting trapped in local optima, enhances the robustness and adaptability of the algorithm under complex working conditions, and enables the system to automatically adjust to the optimal control state according to changes in the working environment.

[0038] Third, by establishing a system dynamic model and simulation optimization platform, seamless integration of offline parameter tuning and online control is achieved, which greatly reduces the dependence of parameter tuning on professional experience, improves the efficiency of engineering applications, and avoids the risks and costs of direct debugging on physical equipment.

[0039] Fourth, by integrating flow sensing, positioning, and wireless transmission modules, a pesticide application system integrating precise control, intelligent monitoring, and remote management has been constructed. This system not only achieves precise control of pesticide flow but also provides data support for the visual management and intelligent decision-making of farmland operations, thus powerfully promoting the technological development of precision agriculture. Attached Figure Description

[0040] Figure 1 This is a block diagram of the overall hardware structure of the variable drug administration control system provided in this embodiment of the invention.

[0041] Figure 2 This is a flowchart of the improved K-means clustering optimization algorithm in this embodiment of the invention.

[0042] Figure 3 This is a schematic diagram illustrating the principle of PID parameter tuning based on an improved algorithm in an embodiment of the present invention.

[0043] Figure 4 This is a flowchart of the main software program of the variable drug administration system according to an embodiment of the present invention. Detailed Implementation

[0044] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0045] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0046] Reference Figure 1 - Figure 4 The variable application system based on IKO-PID parameter tuning disclosed in this invention is mainly applied in the field of agricultural plant protection machinery, and is particularly suitable for variable application scenarios that require precise control of pesticide flow, such as field crops, orchards and greenhouses.

[0047] The variable dosing system based on IKO-PID parameter tuning includes a core controller, an execution module, a signal acquisition module, and a power supply module.

[0048] The core controller uses an embedded microprocessor (such as the STM32 series), with built-in PID control program and improved K-means clustering optimization algorithm (IKO).

[0049] The execution module includes a rotating drum nozzle drive unit and a brushless water pump drive unit. The rotating drum nozzle drive unit controls the rotation speed of the drum through a brushless motor, and the brushless water pump drive unit controls the liquid delivery pressure through a brushless water pump.

[0050] The signal acquisition module includes a flow sensor (such as a Hall effect flow sensor) and a positioning module (such as a GPS / BeiDou module) to collect flow and location information in real time. The power module is a lithium battery pack that provides a stable voltage to each module through a DC-DC converter circuit.

[0051] In a preferred embodiment, the core controller controls the brushless motor in the rotary nozzle drive unit and the brushless water pump in the brushless water pump drive unit respectively by outputting PWM signals. The duty cycle of the PWM signal is calculated in real time by a PID algorithm and adjusted according to the optimal parameters tuned by IKO to ensure that the liquid flow rate accurately tracks the set value.

[0052] The main program flow of the system software includes steps such as initialization, parameter reading, signal acquisition, PID calculation, PWM output and communication processing to achieve closed-loop control.

[0053] In a preferred embodiment, the PID parameter tuning method includes the following steps: First, establishing a dynamic mathematical model of the variable drug delivery system or its actuator motor.

[0054] Secondly, in the simulation environment, the performance indicators of the control system (such as ISE and ITSE) are used as fitness functions, and the IKO algorithm is used to automatically optimize the PID parameters.

[0055] Finally, the optimal PID parameters are written into the core controller.

[0056] The optimization process of the IKO algorithm includes population initialization based on the set of best points, population update incorporating differential evolution strategy, global exploration or local development selection based on location density probability, dynamic location guidance based on the global optimal solution, and optimal solution update based on lens imaging back learning.

[0057] In a preferred embodiment, the improved K-means clustering optimization algorithm (IKO) enhances optimization performance through four strategies: initialization of the optimal point set ensures uniform population distribution.

[0058] Differential evolution operations enhance population diversity; lens imaging reverse learning helps escape local optima.

[0059] Dynamic guidance methods accelerate convergence. The algorithm runs in a simulation platform (such as MATLAB / Simulink), outputs the optimal PID parameter combination, and then writes it into the core controller's Flash memory using a programming tool.

[0060] like Figure 1 As shown:

[0061] The power supply provides power to the entire system.

[0062] The core controller is the brain, receiving feedback from sensors and issuing commands to the actuators.

[0063] The wireless module interacts with the controller to achieve remote communication.

[0064] The entire process forms a typical closed-loop automatic control system of "sensing-decision-execution".

[0065] like Figure 2 As shown, the four key flows of power supply, PWM control, signal feedback and wireless data are clearly marked, intuitively demonstrating closed-loop control. The four innovative strategies (optimal point set, differential evolution, dynamic guidance and lens imaging) are embedded as core steps, clearly showing how they work together to achieve global optimization.

[0066] like Figure 3As shown, offline tuning: the improved optimization algorithm (i.e., the IKO algorithm) runs automatically in the simulation environment, and its sole purpose is to output a set of optimal PID parameters.

[0067] Online control: The obtained optimal parameters are loaded into the PID controller, and the system enters real-time operation.

[0068] The controller, the controlled object (such as a motor or water pump), and the sensor together form a closed loop to achieve precise automatic control.

[0069] Through advanced intelligent algorithms, complex and time-consuming parameter optimization is completed in a virtual environment, and the optimal solution is then seamlessly applied to the physical system, thus solving the pain points of difficult parameter tuning and poor performance of traditional PID control once and for all.

[0070] like Figure 4 As shown, after initialization, the program enters an infinite main loop. In each loop cycle, the system first collects sensor data from the field (such as flow rate and location). The core controller performs PID calculations based on the collected data and target values, converts the calculation results into PWM signals, drives the actuators (motors, water pumps) to move, and finally sends the system status out through the wireless module. This process is repeated to achieve continuous and precise control.

[0071] The automated task of variable drug administration is completed through a cycle of "perception-thinking-execution-communication".

[0072] Working Principle: Upon system power-on, the core controller loads preset PID parameters and the IKO optimization program. The signal acquisition module collects real-time information on pesticide flow rate and application location, transmitting it wirelessly to a remote cloud platform. Based on the deviation between the set and measured flow rates, the core controller uses the PID algorithm to calculate the PWM output value, driving the execution module to adjust the pesticide atomization and delivery process. When system conditions change, the IKO algorithm can readjust the PID parameters online or offline to ensure the system is always in optimal control. Through this method, the system achieves rapid, accurate, and stable control of pesticide flow rate, fundamentally solving the problems of over-spraying and under-spraying.

[0073] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. The substitutions may be replacements of some structures, devices, or method steps, or they may be complete technical solutions. Equivalent substitutions or modifications made to the technical solutions and inventive concepts of the present invention should all be covered within the scope of protection of the present invention.

Claims

1. A variable-rate dosing system based on IKO-PID parameter tuning, characterized in that, include: Core controller, execution module, signal acquisition module, and power supply module; The core controller is used to run the control program and PID control algorithm, and output control signals to the execution module; The execution module includes a rotary nozzle drive unit and a brushless water pump drive unit, used to perform atomization and delivery of the liquid medicine; The signal acquisition module includes a flow sensor, which collects the flow rate of the liquid medicine in real time and feeds it back to the core controller to form a closed-loop control. The power module supplies power to all modules of the system; The parameters of the PID control algorithm are obtained by automatic optimization using an improved K-means clustering optimization algorithm. The improvements include a population initialization strategy, a population diversity enhancement strategy, a local optimum escape strategy, and a convergence acceleration strategy. The automatic optimization process sequentially executes population initialization based on the set of best points, population update incorporating differential evolution strategy, global exploration or local development selection based on location density probability, dynamic location guidance based on the global optimal solution, and optimal solution update based on lens imaging reverse learning. The population initialization strategy is to generate an initial population using the optimal point set theory, so that the population is evenly distributed in the parameter search space. The population diversity enhancement strategy is to introduce differential evolution operations during the algorithm iteration process; The local optimal escape strategy is to use a lens imaging back learning method to process the current optimal solution in order to explore potential better solution space regions. The convergence acceleration strategy is to adopt a dynamic guidance method based on the global historical optimal solution, so that the search agent moves towards the optimal solution during the iteration process; The core controller outputs PWM signals to control the brushless motor in the rotary nozzle drive unit and the brushless water pump in the brushless water pump drive unit, respectively.

2. The variable dosing system based on IKO-PID parameter tuning according to claim 1, characterized in that, The signal acquisition module also includes a positioning module, and the system also includes a wireless transmission module, which interacts with and remotely controls a remote cloud platform.

3. A method for tuning PID parameters in a variable-rate dosing system according to any one of claims 1-2, characterized in that, Includes the following steps: S1: Establish a dynamic mathematical model of the variable drug delivery system or its actuator motor based on the IKO-PID parameter tuning; S2: In the simulation environment, the performance index of the control system of the model is used as the fitness function, and the improved K-means clustering optimization algorithm is used to automatically optimize the proportional, integral and derivative parameters of the PID controller. S3: After the iterative optimization of the variable dosing system algorithm based on IKO-PID parameter tuning is completed, the optimal combination of PID parameters is output and written into the control program of the core controller; The automatic optimization process described in S2 executes the following steps in sequence: population initialization based on the set of best points, population update incorporating differential evolution strategy, global exploration or local development selection based on location density probability, dynamic location guidance based on the global optimal solution, and optimal solution update based on lens imaging reverse learning.

4. The PID parameter tuning method according to claim 3, characterized in that, The performance indicators of the control system in S2 are expressed as integral squared error (ISE) or time-weighted integral squared error (ITSE).

Images