Self-adaptive material accurate quantitative distribution method and system and storage medium

Abstract

The invention provides a self-adaptive material precise quantitative distribution method and system and a storage medium, and the method comprises the steps: obtaining the tracking error and change rate of a target quantification and a current cumulant, estimating the system state and total disturbance in real time through a fractional order expansion state observer, decomposing the total disturbance into instantaneous and slowly varying components, and obtaining the target quantification and current cumulant. Adjusting the fractional order of the observer on line according to the amplitude ratio of the two, correcting an overshoot prediction model by using a slowly varying component, determining a distribution stop target value, constructing a fractional order non-linear state error feedback control law in a control link, adjusting a control gain according to slowly varying disturbance, and adjusting a non-linear power according to an accumulation progress. And a control signal is generated in combination with instantaneous disturbance compensation until the material accumulation amount reaches a stop target value, and quantitative distribution is achieved.

Description

Technical Field

[0001] This application belongs to the field of control, and in particular relates to an adaptive method, system and storage medium for precise quantitative distribution of materials. Background Technology

[0002] Material distribution systems typically consist of storage bins, actuators, and weighing sensors. The goal is to ensure that the accumulated material quantity is as close as possible to a preset target quantity while meeting speed requirements. In actual distribution, various internal and external disturbances are unavoidable, such as fluctuations in material properties, nonlinear characteristics and time-varying wear of the actuators, environmental vibrations, and voltage fluctuations. The delay between the actuator stopping and the material completely falling into the weighing unit, known as overshoot or air gap, is one of the main causes of distribution errors. Traditional control methods, such as proportional-integral-derivative (PID) control, often rely on experience for parameter tuning, making it difficult to balance speed and accuracy. Active disturbance rejection (ADRC) technology enhances the system's disturbance rejection capability by using an extended state observer to uniformly estimate all uncertainties and disturbances within and outside the system as a total disturbance and perform real-time compensation. However, its ability to estimate disturbances containing both high-frequency and low-frequency components, or those with fractional-order characteristics, is limited. For overshoot problems, existing prediction methods often use offline calibrated empirical formulas or linear models, which cannot be adjusted online when material properties or operating conditions change, affecting distribution accuracy. Therefore, how to achieve an allocation method that can simultaneously and accurately estimate and compensate for complex disturbances, predict overshoot online, and optimize control strategies is a technical problem that urgently needs to be solved in the current field. Summary of the Invention

[0003] This invention proposes an adaptive method for precise quantitative material distribution to address the problem that existing technologies struggle to accurately estimate and compensate for complex disturbances and cannot predict overshoot online. The method includes the following steps:

[0004] The system obtains the initial target quantitative value of material allocation and the current cumulative material quantity collected in real time, calculates the tracking error and the rate of change of tracking error between the two, and estimates the system state and total disturbance in real time using the tracking error and the rate of change of tracking error through a fractional extended state observer.

[0005] The estimated total disturbance is decomposed into instantaneous disturbance components and slowly varying disturbance components. Based on the amplitude ratio of the instantaneous disturbance components and the slowly varying disturbance components, the fractional order of the fractional extended state observer is adjusted online. The overshoot prediction model is corrected online using the slowly varying disturbance components, and the assigned stopping target value is determined based on the corrected model.

[0006] A fractional-order nonlinear state error feedback control law is constructed. The basic gain of the nonlinear function in the control law is adjusted online according to the slowly varying disturbance component. The power coefficient of the nonlinear function is adjusted according to the ratio of the current material accumulation to the initial target quantitative value. The feedback control quantity generated based on the system state estimate and the adjusted nonlinear function is combined with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material distribution actuator.

[0007] The control signal is continuously output until the current accumulated amount of material reaches the distribution stop target value.

[0008] Optionally, the step of decomposing the estimated total disturbance into instantaneous disturbance components and slowly varying disturbance components, and adjusting the fractional order of the fractional-order extended state observer online according to the amplitude ratio of the instantaneous disturbance components and the slowly varying disturbance components, includes:

[0009] The slowly varying disturbance component in the estimate of the total disturbance is obtained by using a low-pass filter.

[0010] The instantaneous disturbance component is obtained by subtracting the estimated value of the total disturbance from the slowly varying disturbance component.

[0011] Calculate the amplitude ratio of the instantaneous disturbance component to the slowly varying disturbance component, and adjust the fractional order online based on the ratio.

[0012] Optionally, the step of using the slowly varying disturbance component to online correct the overshoot prediction model and determining the assigned stopping target value based on the corrected model includes:

[0013] Construct an overshoot prediction model;

[0014] The corrected model is used to determine the target value for assignment stopping.

[0015] Optionally, the step of adjusting the base gain of the nonlinear function in the control law online according to the slowly varying disturbance component includes:

[0016] The base gain is adjusted online based on the slowly varying disturbance component using a preset function, which causes the base gain to decrease as the amplitude of the slowly varying disturbance component increases.

[0017] Optionally, adjusting the power coefficient of the nonlinear function according to the ratio of the current accumulated material quantity to the initial target quantitative value includes:

[0018] When the ratio is lower than a preset threshold, the power coefficient is set to the first coefficient value;

[0019] When the ratio is higher than the preset threshold, the power coefficient is reduced from the first coefficient value to the second coefficient value, and the first coefficient value is greater than the second coefficient value.

[0020] Optionally, the step of combining the feedback control quantity generated based on the system state estimate and the adjusted nonlinear function with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material distribution actuator includes:

[0021] Based on the system state estimate, the nonlinear feedback control quantity is calculated using the adjusted base gain and the adjusted power coefficient.

[0022] Calculate the compensation amount used to counteract the instantaneous disturbance component;

[0023] The control signal is generated by subtracting the compensation amount from the nonlinear feedback control amount.

[0024] Furthermore, this invention also proposes an adaptive material precise quantitative distribution system, comprising the following modules:

[0025] The calculation module is used to obtain the initial target quantitative value of material allocation and the current cumulative amount of material collected in real time, calculate the tracking error and the rate of change of tracking error between the two; and use the tracking error and the rate of change of tracking error to estimate the system state and total disturbance in real time through the fractional extended state observer.

[0026] The determination module is used to decompose the estimated value of the total disturbance into instantaneous disturbance components and slowly varying disturbance components, and adjust the fractional order of the fractional-order extended state observer online according to the amplitude ratio of the instantaneous disturbance components and the slowly varying disturbance components; use the slowly varying disturbance components to correct the overshoot prediction model online, and determine the allocation stop target value based on the corrected model;

[0027] The generation module is used to construct a fractional-order nonlinear state error feedback control law, adjust the base gain of the nonlinear function in the control law online according to the slowly varying disturbance component, and adjust the power coefficient of the nonlinear function according to the ratio of the current material accumulation to the initial target quantitative value; and combine the feedback control quantity generated based on the system state estimate and the adjusted nonlinear function with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material distribution actuator.

[0028] The output module is used to continuously output the control signal until the current material accumulation reaches the distribution stop target value.

[0029] Preferably, the step of decomposing the estimated total disturbance into instantaneous disturbance components and slowly varying disturbance components, and adjusting the fractional order of the fractional-order extended state observer online according to the amplitude ratio of the instantaneous disturbance components and the slowly varying disturbance components, includes:

[0030] The slowly varying disturbance component in the estimate of the total disturbance is obtained by using a low-pass filter.

[0031] The instantaneous disturbance component is obtained by subtracting the estimated value of the total disturbance from the slowly varying disturbance component.

[0032] Calculate the amplitude ratio of the instantaneous disturbance component to the slowly varying disturbance component, and adjust the fractional order online based on the ratio.

[0033] Preferably, the step of using the slowly varying disturbance component to online correct the overshoot prediction model and determining the assigned stopping target value based on the corrected model includes:

[0034] Construct an overshoot prediction model;

[0035] The corrected model is used to determine the target value for assignment stopping.

[0036] Preferably, the step of adjusting the base gain of the nonlinear function in the control law online according to the slowly varying disturbance component includes:

[0037] The base gain is adjusted online based on the slowly varying disturbance component using a preset function, which causes the base gain to decrease as the amplitude of the slowly varying disturbance component increases.

[0038] Preferably, adjusting the power coefficient of the nonlinear function according to the ratio of the current accumulated material quantity to the initial target quantitative value includes:

[0039] When the ratio is lower than a preset threshold, the power coefficient is set to the first coefficient value;

[0040] When the ratio is higher than the preset threshold, the power coefficient is reduced from the first coefficient value to the second coefficient value, and the first coefficient value is greater than the second coefficient value.

[0041] Preferably, the control signal for the material distribution actuator, which combines the feedback control quantity generated based on the system state estimate and the adjusted nonlinear function with the compensation quantity for the instantaneous disturbance component, includes:

[0042] Based on the system state estimate, the nonlinear feedback control quantity is calculated using the adjusted base gain and the adjusted power coefficient.

[0043] Calculate the compensation amount used to counteract the instantaneous disturbance component;

[0044] The control signal is generated by subtracting the compensation amount from the nonlinear feedback control amount.

[0045] This invention enables rapid and accurate estimation of total disturbances during system operation, identifying both instantaneous strong disturbances and slowly varying disturbance components. This allows for optimization of the observer's performance based on different disturbance characteristics, improving the reliability of state estimation. By utilizing slowly varying disturbance information to perform real-time correction of the overshoot prediction model, the prediction accuracy of the allocation stop target value is improved, suppressing deviations in quantitative allocation. During control, key parameters of the nonlinear feedback control law are adjusted based on the slowly varying disturbances and the current allocation progress, balancing rapid response in the initial allocation phase with a smooth transition as the target approaches, achieving a good combination of allocation speed and quantitative accuracy. The compensation mechanism for instantaneous disturbances further enhances the system's ability to resist sudden interference, ensuring the stability of the allocation process and improving the reliability of quantitative material allocation. Attached Figure Description

[0046] Figure 1 A flowchart of the first embodiment;

[0047] Figure 2 A schematic diagram for estimating the system state and total disturbance;

[0048] Figure 3 A schematic diagram of the model for predicting overshoot and determining the stop target value;

[0049] Figure 4 This is a schematic diagram of an adaptive material precise quantitative distribution system control. Detailed Implementation

[0050] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0051] See the first embodiment. Figure 1 An adaptive method for precise quantitative distribution of materials includes the following steps:

[0052] S1, obtain the initial target quantitative value of material allocation and the current cumulative amount of material collected in real time, calculate the tracking error and the rate of change of tracking error between the two; through the fractional extended state observer, use the tracking error and the rate of change of tracking error to estimate the system state and total disturbance in real time;

[0053] The initial target quantity for material distribution, such as 1000g, is set and stored through a human-machine interface. A weighing sensor and data acquisition module installed below the silo collect the current material weight, thus obtaining the current cumulative material quantity. Subtracting the current cumulative material quantity from the initial target quantity yields the tracking error at the current moment. Subtracting the tracking error from the previous moment from the current tracking error, and then dividing by the sampling period, gives the rate of change of the tracking error.

[0054] Establish a third-order fractional extended state observer, with state variables... , , These are used to estimate the tracking error, the rate of change of the tracking error, and the total disturbance, respectively. Using the state estimate from the previous time step and the tracking error at the current time step, and through a preset observer gain and a fractional-order α, the state estimate at the current time step is iteratively calculated according to the discretized difference equation of the fractional-order extended state observer. , The estimated value of total disturbance .

[0055] In an optional embodiment, the step of estimating the system state and total disturbance in real time using the tracking error and the rate of change of the tracking error via a fractional-order extended state observer includes:

[0056] Real-time estimation of the system state and total disturbance is achieved by establishing the following fractional-order extended state observer state equations:

[0057] set up This is an estimate of the tracking error e(t). To track the rate of change of error The estimated value, Given an estimate of the total disturbance, the state equation is:

[0058]

[0059] in, express Fractional differential operators of order, Let u(t) be the fractional order, and u(t) be the control signal. To control the gain. For observer bandwidth parameters, This is the preset dimensionless observer gain coefficient.

[0060] Initialize the observer parameters. For example, set the fractional order. The initial value is 0.9, and the control gain is... Set the observer bandwidth to 25 based on system characteristics. and take The observer gains are calculated as follows: , , At the same time, the observer state variable , , The initial values ​​are all set to 0.

[0061] See Figure 2 State estimation is performed within each control cycle. At time t, the actual cumulative material quantity is collected, and the difference between this and the expected cumulative quantity trajectory is used to obtain the tracking error e(t). This error e(t) and the control signal u(t) output from the previous cycle are used as observer inputs. Based on the definition of Grünwald-Letnikov fractional differential, the above continuous state equation is discretized into a fractional difference equation for numerical iterative solution. Update At that time, e(t) will be used in conjunction with the previous time step. The difference, combined with the gain And the previous moment Value, calculate the current time. Similarly, update and After calculation, the observer outputs the estimated system state at the current moment. , Total disturbance estimate ,For example The values ​​will be used for subsequent controller planning and disturbance compensation. The observer model is structured as a third-order state observer, extended by one state. It actively estimates and observes the total disturbances inside and outside the system, and improves the ability to represent complex processes through fractional-order operators.

[0062] S2, decompose the estimated value of the total disturbance into instantaneous disturbance components and slowly varying disturbance components, and adjust the fractional order of the fractional-order extended state observer online according to the amplitude ratio of the instantaneous disturbance components and the slowly varying disturbance components; use the slowly varying disturbance components to correct the overshoot prediction model online, and determine the allocation stop target value based on the corrected model;

[0063] Specifically, the estimated value of the total disturbance Inputting a first-order low-pass filter, the output of the filter is the slowly varying disturbance component, and the total disturbance estimate is... Subtracting the slowly varying perturbation component yields the instantaneous perturbation component. Calculate the root mean square (RMS) value of the instantaneous perturbation component within a sliding time window, and its ratio to the RMS value of the slowly varying perturbation component within the same time window. Set upper and lower limits for the fractional order α, for example, between 0.8 and 1.0. Based on the magnitude of this ratio, adjust the value of order α online through linear mapping or a preset functional relationship.

[0064] After each task allocation, the actual measured overshoot is compared with the model-predicted overshoot. The difference is combined with the mean of the slowly varying perturbation components obtained in the previous batch as a correction signal. The model parameters are then updated online using algorithms such as least squares or gradient descent to correct model drift. The initial target quantitative value of 1000g is subtracted from the predicted overshoot calculated based on the corrected model to obtain the allocation stopping target value, for example, 995g.

[0065] In an optional embodiment, the step of decomposing the estimated total disturbance into instantaneous disturbance components and slowly varying disturbance components, and adjusting the fractional order of the fractional-order extended state observer online according to the amplitude ratio of the instantaneous disturbance components to the slowly varying disturbance components, includes:

[0066] The slowly varying disturbance component in the estimate of the total disturbance is obtained by using a low-pass filter.

[0067] The instantaneous disturbance component is obtained by subtracting the estimated value of the total disturbance from the slowly varying disturbance component.

[0068] Calculate the amplitude ratio of the instantaneous disturbance component to the slowly varying disturbance component, and adjust the fractional order online based on the ratio.

[0069] The previous step yielded the total disturbance estimate. The input is fed into a first-order low-pass filter, and the filter's transfer function is... , where s is the Laplace operator and the time constant. It can be set to 0.1 seconds. The filter output is the slowly varying disturbance component. Subtracting the slowly varying disturbance component from the total disturbance estimate yields the instantaneous disturbance component. For example, if at a certain moment The value is 50.3, obtained after filtering. If the value is 45.1, then the instantaneous disturbance component It is 5.2.

[0070] To assess the relative strength of the two types of disturbances, their root mean square values ​​over a time window, such as the most recent 50 sampling points, are calculated as their respective amplitudes. and Calculate the amplitude ratio. In one embodiment, the fractional order for ,in , , , When the proportion of instantaneous disturbances is large, i.e., the R value is large, It will approach 1.0, making the observer closer to an integer order form to enhance its ability to smooth high-frequency noise; when slowly varying disturbances dominate, i.e., when the R value is small, It will approach 0.8, utilizing the memory properties of fractional operators to improve the tracking accuracy of slowly changing processes. (Adjusted) The value will be used to calculate the fractional-order extended state observer at the next time step.

[0071] In an optional embodiment, the step of using the slowly varying disturbance component to online correct the overshoot prediction model and determining the assigned stopping target value based on the corrected model includes:

[0072] Construct an overshoot prediction model;

[0073] The corrected model is used to determine the target value for assignment stopping.

[0074] The impulse prediction model is preferably a linear superposition structure: Where v(t) is the current feed rate, and k and c are the baseline model parameters calibrated using historical data. For slowly varying disturbances, To obtain the slowly varying perturbation components. For example, k=0.05s, c=1.0g can be calibrated experimentally. =0.08g / perturbation unit.

[0075] Assume the initial target quantity is 500g. During the final stage of feeding, continuous monitoring is conducted. If at a certain moment, v(t) = 100 g / s is detected, This indicates the existence of a persistent disturbance that increases the feed rate. In this case, the expected overshoot is calculated using a predictive model: g. Adjust the allocation stop target value to: g. The controller will use this adjusted value as the trigger point for the stop command, closing the feed valve in advance, thereby using the predicted overshoot to ensure the final weight reaches the target value.

[0076] S3, construct a fractional-order nonlinear state error feedback control law, adjust the basic gain of the nonlinear function in the control law online according to the slowly varying disturbance component, and adjust the power coefficient of the nonlinear function according to the ratio of the current material accumulation to the initial target quantitative value; combine the feedback control quantity generated based on the system state estimate and the adjusted nonlinear function with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material distribution actuator;

[0077] Specifically, nonlinear functions are used. Set the base gain The adjustment rule is that when the absolute value of the slowly varying disturbance component increases, it is reduced according to a preset function. The value of the power coefficient 'a' is set as follows: as the ratio of the current accumulated material quantity to the initial target quantitative value gradually increases from 0 to 1, the power coefficient smoothly decreases from a large value to a small value, thereby achieving rapid feeding in the initial stage and fine control in the final stage. Feedback control quantity. Equal to the adjusted base gain The control signal u is equal to the feedback control quantity. The compensation amount is subtracted from the instantaneous disturbance component.

[0078] In an optional embodiment, the step of adjusting the base gain of the nonlinear function in the control law online according to the slowly varying disturbance component includes:

[0079] The base gain is adjusted online based on the slowly varying disturbance component using a preset function, which causes the base gain to decrease as the amplitude of the slowly varying disturbance component increases.

[0080] An example function is .in, The maximum base gain, such as 150, is used when the system is stable and undisturbed to ensure a fast response; The minimum base gain, for example, 40, serves as a lower limit to prevent the system response from being too sluggish; This is the attenuation coefficient, for example, 0.1, used to adjust the sensitivity of the gain to changes in disturbance; This represents the absolute value of the current slowly varying disturbance component.

[0081] See Figure 3 In the control loop, the disturbance decomposition module is used to obtain... Suppose that at a certain stage, The absolute value of 25 indicates the presence of a persistent disturbance. The controller calculates the current base gain based on a preset function: The reduced base gain The original fixed gain will be replaced by a variable gain used to calculate subsequent nonlinear feedback control values. In this way, when the system is affected by slowly varying disturbances such as continuous fluctuations in pipeline pressure, the controller automatically becomes more conservative, suppressing oscillations or instability that may be caused by excessively high gain and disturbances.

[0082] In an optional embodiment, adjusting the power coefficient of the nonlinear function according to the ratio of the current material accumulation to the initial target quantitative value includes:

[0083] When the ratio is lower than a preset threshold, the power coefficient is set to the first coefficient value;

[0084] When the ratio is higher than the preset threshold, the power coefficient is reduced from the first coefficient value to the second coefficient value, and the first coefficient value is greater than the second coefficient value.

[0085] Assume the initial target quantification is 1000g. Set a proportion threshold of 0.85. Set the first coefficient value. A value of 1.2, a relatively large value, enables the controller to exert a stronger control effect when the error is large, achieving rapid feeding. Set the second coefficient value. A value of 0.5, a relatively small value, allows the controller to operate more gently when the error is small, reducing the impact when approaching the target value.

[0086] The power factor is adjusted in stages during the feeding process. At the initial stage of feeding, the current accumulated material amount is monitored. .if only That is, the cumulative amount has not reached 850g, power factor The initial coefficient value is kept constant at 1.2. When the accumulated amount reaches or exceeds 850g, a decreasing adjustment mechanism is activated. (Power factor) Linear interpolation will be performed based on the current cumulative amount's position within the 850g to 1000g range: For example, when the cumulative amount is 925g, The adjustment mentioned The value will be used to calculate the nonlinear feedback control quantity, realizing a smooth transition from the initial large opening and rapid feeding to the later small opening control.

[0087] In an optional embodiment, the step of combining the feedback control quantity generated based on the system state estimate and the adjusted nonlinear function with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material dispensing actuator includes:

[0088] Based on the system state estimate, the nonlinear feedback control quantity is calculated using the adjusted base gain and the adjusted power coefficient.

[0089] Calculate the compensation amount used to counteract the instantaneous disturbance component;

[0090] The control signal is generated by subtracting the compensation amount from the nonlinear feedback control amount.

[0091] Nonlinear feedback control quantity Based on the nonlinear proportional-differential structure, the formula is: .in, , The tracking error and the rate of change of tracking error are estimated for the state observer; It is the base gain after online adjustment based on the slowly varying disturbance; a(t) is the power coefficient adjusted according to the feed ratio; It is a fixed differential gain. (Function) It is a saturation function, when Time equals Otherwise Suppose at this moment... , , , , ,but The calculation results reflect the main control effect on the current system deviation and trend.

[0092] Feedforward compensation Specifically designed to counteract transient disturbances in a system, the calculation formula is as follows: .in, The instantaneous disturbance component separated from the total disturbance, such as the impact caused by the falling of material clumps; The known system control gain. If at this moment, it is detected... ,and The compensation amount is The magnitude of this compensation corresponds to the control force required to counteract the instantaneous impact.

[0093] The nonlinear feedback control quantity is combined with the feedforward compensation quantity, and the control signal is: By subtraction, the controller, while performing feedback control based on system state, actively and inversely applies a control force equivalent to the impact of the instantaneous disturbance, thereby achieving rapid suppression of sudden disturbances. This synthesized control signal u(t) is sent to the actuator, such as the driver of a feed valve, to control the material flow rate. Figure 4 .

[0094] S4, continuously output the control signal until the current material accumulation reaches the distribution stop target value.

[0095] Specifically, the calculated control signal u is converted into a standard 4-20mA current signal or 0-10V voltage signal via a digital-to-analog converter, and output to the frequency converter that controls the speed of the screw feeder. The controller cyclically executes all the above calculation and control steps within each sampling cycle, and compares the current accumulated material amount collected by the weighing sensor with the distribution stop target value of 995g in real time. When the current accumulated material amount is greater than or equal to 995g, the control signal output to the frequency converter is immediately set to zero, causing the screw feeder to stop rotating, completing this quantitative distribution.

[0096] In a second embodiment, the present invention also provides an adaptive material precise quantitative distribution system, comprising the following modules:

[0097] The calculation module is used to obtain the initial target quantitative value of material allocation and the current cumulative amount of material collected in real time, calculate the tracking error and the rate of change of tracking error between the two; and use the tracking error and the rate of change of tracking error to estimate the system state and total disturbance in real time through the fractional extended state observer.

[0098] The determination module is used to decompose the estimated value of the total disturbance into instantaneous disturbance components and slowly varying disturbance components, and adjust the fractional order of the fractional-order extended state observer online according to the amplitude ratio of the instantaneous disturbance components and the slowly varying disturbance components; use the slowly varying disturbance components to correct the overshoot prediction model online, and determine the allocation stop target value based on the corrected model;

[0099] The generation module is used to construct a fractional-order nonlinear state error feedback control law, adjust the base gain of the nonlinear function in the control law online according to the slowly varying disturbance component, and adjust the power coefficient of the nonlinear function according to the ratio of the current material accumulation to the initial target quantitative value; and combine the feedback control quantity generated based on the system state estimate and the adjusted nonlinear function with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material distribution actuator.

[0100] The output module is used to continuously output the control signal until the current material accumulation reaches the distribution stop target value.

[0101] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0102] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An adaptive method for precise quantitative distribution of materials, characterized in that, Includes the following steps: Obtain the initial target quantitative value for material allocation and the current cumulative material quantity collected in real time, and calculate the tracking error and the rate of change of the tracking error between the two; By using a fractional-order extended state observer, the system state and total disturbance are estimated in real time using the tracking error and the rate of change of the tracking error. The estimated total disturbance is decomposed into instantaneous disturbance components and slowly varying disturbance components. Based on the amplitude ratio of the instantaneous disturbance components and the slowly varying disturbance components, the fractional order of the fractional extended state observer is adjusted online. The overshoot prediction model is corrected online using the slowly varying disturbance components, and the assigned stopping target value is determined based on the corrected model. A fractional-order nonlinear state error feedback control law is constructed. The basic gain of the nonlinear function in the control law is adjusted online according to the slowly varying disturbance component. The power coefficient of the nonlinear function is adjusted according to the ratio of the current material accumulation to the initial target quantitative value. The feedback control quantity generated based on the system state estimate and the adjusted nonlinear function is combined with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material distribution actuator. The control signal is continuously output until the current accumulated amount of material reaches the distribution stop target value.

2. The method according to claim 1, characterized in that, The step of decomposing the estimated total disturbance into instantaneous disturbance components and slowly varying disturbance components includes: The slowly varying disturbance component in the estimate of the total disturbance is obtained by using a low-pass filter. The instantaneous disturbance component is obtained by subtracting the estimated total disturbance from the slowly varying disturbance component.

3. The method according to claim 1 or 2, characterized in that, The step of adjusting the base gain of the nonlinear function in the control law online according to the slowly varying disturbance component includes: The base gain is adjusted online based on the slowly varying disturbance component using a preset function, which causes the base gain to decrease as the amplitude of the slowly varying disturbance component increases.

4. The method according to claim 1, characterized in that, The step of adjusting the power coefficient of the nonlinear function according to the ratio of the current accumulated material quantity to the initial target quantitative value includes: When the ratio is lower than a preset threshold, the power coefficient is set to the first coefficient value; When the ratio is higher than the preset threshold, the power coefficient is reduced from the first coefficient value to the second coefficient value, and the first coefficient value is greater than the second coefficient value.

5. The method according to claim 4, characterized in that, The step of combining the feedback control quantity generated based on the system state estimate and the adjusted nonlinear function with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material distribution actuator includes: Based on the system state estimate, the nonlinear feedback control quantity is calculated using the adjusted base gain and the adjusted power coefficient. Calculate the compensation amount used to counteract the instantaneous disturbance component; The control signal is generated by subtracting the compensation amount from the nonlinear feedback control amount.

6. An adaptive material precise quantitative distribution system, characterized in that, Includes the following modules: The calculation module is used to obtain the initial target quantitative value of material allocation and the current cumulative amount of material collected in real time, and to calculate the tracking error and the rate of change of tracking error between the two. By using a fractional-order extended state observer, the system state and total disturbance are estimated in real time using the tracking error and the rate of change of the tracking error. The determination module is used to decompose the estimated value of the total disturbance into instantaneous disturbance components and slowly varying disturbance components, and adjust the fractional order of the fractional-order extended state observer online according to the amplitude ratio of the instantaneous disturbance components and the slowly varying disturbance components; use the slowly varying disturbance components to correct the overshoot prediction model online, and determine the allocation stop target value based on the corrected model; The generation module is used to construct a fractional-order nonlinear state error feedback control law, adjust the base gain of the nonlinear function in the control law online according to the slowly varying disturbance component, and adjust the power coefficient of the nonlinear function according to the ratio of the current material accumulation to the initial target quantitative value; and combine the feedback control quantity generated based on the system state estimate and the adjusted nonlinear function with the compensation quantity for the instantaneous disturbance component to generate a control signal for the material distribution actuator. The output module is used to continuously output the control signal until the current material accumulation reaches the distribution stop target value.

7. The system according to claim 6, characterized in that, The step of decomposing the estimated total disturbance into instantaneous disturbance components and slowly varying disturbance components includes: The slowly varying disturbance component in the estimate of the total disturbance is obtained by using a low-pass filter. The instantaneous disturbance component is obtained by subtracting the estimated total disturbance from the slowly varying disturbance component.

8. The system according to claim 6, characterized in that, The step of adjusting the base gain of the nonlinear function in the control law online according to the slowly varying disturbance component includes: The base gain is adjusted online based on the slowly varying disturbance component using a preset function, which causes the base gain to decrease as the amplitude of the slowly varying disturbance component increases.

9. The system according to claim 6, characterized in that, The step of adjusting the power coefficient of the nonlinear function according to the ratio of the current accumulated material quantity to the initial target quantitative value includes: When the ratio is lower than a preset threshold, the power coefficient is set to the first coefficient value; When the ratio is higher than the preset threshold, the power coefficient is reduced from the first coefficient value to the second coefficient value, and the first coefficient value is greater than the second coefficient value.

10. A computer-readable storage medium storing a computer program thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-5.

Images