A quality control system and method for a birch water production line

Abstract

This invention relates to the field of industrial automation control technology, and particularly to a quality control system and method for a sparkling water production line. The system includes: acquiring physical quantities of the control process, constructing a state vector, quantifying the deviation of the state vector, constructing a total potential energy function, and obtaining the total system potential energy; constructing a control device action vector, quantifying the changing trend of the control device action vector with respect to the total system potential energy, and obtaining an action gradient vector; introducing a disturbance response mechanism, constructing a disturbance correction vector and a disturbance potential energy function based on the state vector; constructing a control optimization objective function based on the disturbance potential energy function, and dynamically updating the control device action vector through gradient solving and step size control. This solves the problems of traditional quality control methods lacking a unified scheduling mechanism at the system level, lacking the ability to predict and respond to disturbance trends, and relying on empirical parameters or preset rules in the control command generation process, resulting in low control accuracy and poor adjustment efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial automation control, and particularly to a quality control system and method for a bitter apricot bubble water production line. Background Art

[0002] With the increasing demand of consumers for healthy drinks, diverse flavors and high-quality beverages, bitter apricot bubble water, as a new type of drink that combines natural fruit flavors, carbonated refreshing taste and functional formulas, has gradually occupied an important position in the market. Its production process involves multiple links such as fruit pulp ratio, sweetener injection, carbonation saturation control, filling liquid level management, etc. There are complex dynamic correlations among these parameters, and they have a direct impact on the taste and quality stability of the final product.

[0003] To meet the requirements of efficient, high-quality and continuous production, modern beverage manufacturing is gradually developing towards a highly automated and intelligent direction, putting forward higher technical requirements for the quality control system of the production line in terms of real-time performance, coordination and adaptive adjustment ability. Therefore, there is an urgent need to construct a quality control method and system that can achieve multi-parameter linkage control in a dynamic multi-disturbance environment to ensure the taste consistency, filling accuracy and carbonation stability of bitter apricot bubble water, and support the process requirements of its large-scale and standardized production.

[0004] However, traditional quality control methods have the following technical problems: they often adopt an independent and separate strategy, lacking a unified scheduling mechanism at the system level, resulting in mutual interference between control objects and unable to achieve global optimization; lacking the ability to predict and respond to disturbance trends, and it is difficult to meet the production requirements in a complex dynamic environment; the process of generating control instructions depends on empirical parameters or preset rules, and the physical logic is not clear, making it difficult to establish an accurate mapping relationship between control devices and states, resulting in low control accuracy and poor adjustment efficiency, ultimately affecting product consistency and production stability. Summary of the Invention

[0005] The present invention provides a quality control system and method for a bitter apricot bubble water production line to solve the problems that traditional quality control methods often adopt an independent and separate strategy, lacking a unified scheduling mechanism at the system level, resulting in mutual interference between control objects and unable to achieve global optimization; lacking the ability to predict and respond to disturbance trends, and it is difficult to meet the production requirements in a complex dynamic environment; the process of generating control instructions depends on empirical parameters or preset rules, and the physical logic is not clear, making it difficult to establish an accurate mapping relationship between control devices and states, resulting in low control accuracy and poor adjustment efficiency, ultimately affecting product consistency and production stability.

[0006] A quality control system and method for a bitter apricot bubble water production line of the present invention specifically includes the following technical solutions:

[0007] A quality control method for a pistachio sparkling water production line includes the following steps:

[0008] S1. Collect the physical quantities of the control process, construct the state vector, quantify the deviation of the state vector, construct the total potential energy function, and obtain the total potential energy of the system; construct the action vector of the control equipment, quantify the changing trend of the action vector of the control equipment with respect to the total potential energy of the system, and obtain the action gradient vector.

[0009] S2. Introduce a disturbance response mechanism. Based on the state vector, construct a disturbance correction vector and a disturbance potential energy function. Based on the disturbance potential energy function, construct a control optimization objective function and dynamically update the control device action vector through gradient solving and step size control.

[0010] Preferably, S1 specifically includes:

[0011] Introduce the desired state vector and compare the current state vector with the desired state vector dimension by dimension to obtain the state deviation vector; construct the local potential energy function based on the state deviation vector; sum the local potential energy functions to construct the total potential energy function.

[0012] Preferably, S1 specifically includes:

[0013] By introducing a device influence mapping matrix, the relationship between the control device action vector and the state vector is modeled as a static linear mapping.

[0014] Preferably, S1 specifically includes:

[0015] According to the chain rule, the gradient of the total potential energy function with respect to the action vector of the control device is calculated to obtain the action gradient vector.

[0016] Preferably, S2 specifically includes:

[0017] In the implementation of the disturbance response mechanism, the disturbance factor is represented as a time-dependent disturbance vector. Based on the second derivative of the state vector and the first derivative of the time-dependent disturbance vector, a disturbance correction vector is constructed. Based on the disturbance correction vector, a disturbance potential energy function is constructed.

[0018] Preferably, S2 specifically includes:

[0019] The disturbance potential energy function is combined with the total potential energy function to construct the control optimization objective function; the control gradient vector is obtained by calculating the gradient of the control optimization objective function with respect to the control device action vector.

[0020] Preferably, S2 specifically includes:

[0021] Based on the control device action vector at the current moment, and combined with the control gradient vector, the control device action vector for the next moment is generated.

[0022] A quality control system for a pistachio sparkling water production line includes the following components:

[0023] The system includes a state acquisition module, a state modeling module, a motion gradient calculation module, a disturbance response module, and a control command generation module.

[0024] The status acquisition module periodically acquires the physical quantities in the pistachio sparkling water production line and transmits the acquired physical quantities to the status modeling module.

[0025] The state modeling module constructs the current state vector of the system based on physical quantities, and constructs the desired state vector by combining the target formula parameters. By comparing the current state vector with the desired state vector dimension by dimension, it generates a state deviation vector and introduces control weights for the state components to construct a local potential energy function. The local potential energy functions are summed to construct the total potential energy function. The state vector is then passed to the disturbance response module and the action gradient calculation module. Finally, the total potential energy function is passed to the action gradient calculation module.

[0026] The motion gradient calculation module constructs the motion vector of the control device, calculates the sensitivity of the control device's motion vector to changes in total potential energy, and generates the motion gradient vector; it then transmits the control device's motion vector to the disturbance response module and the control command generation module; finally, it transmits the motion gradient vector to the control command generation module.

[0027] The disturbance response module introduces the influence of external disturbance factors into the control logic of the quality control system to obtain a time-dependent disturbance vector; based on the second derivative of the state vector and the first derivative of the time-dependent disturbance vector, a disturbance correction vector is constructed; based on the disturbance correction vector, a disturbance potential energy function is constructed, and the gradient of the disturbance potential energy function with respect to the control device action vector is calculated to generate a disturbance trend response term; the disturbance trend response term is then passed to the control command generation module.

[0028] The control command generation module merges the action gradient vector with the disturbance trend response term to generate the control gradient vector; based on the control device action vector at the current moment, and combined with the control gradient vector, the control device action vector is dynamically updated through gradient solving and step size control.

[0029] The beneficial effects of the technical solution of the present invention are:

[0030] 1. This invention effectively quantifies the impact of state deviation on product quality consistency by constructing a mathematical expression model of local potential energy and total system potential energy, thereby achieving unified optimization control of multi-dimensional parameters; realizing high coordination between control objectives, eliminating mutual interference between control loops, and improving the overall stability and regulation efficiency of the quality control system.

[0031] 2. This invention constructs a linear response model between the action vector of the control device and the state vector of the system, so that the effect of each control device on the control target has a clear quantitative expression; the quality control system can automatically determine the control load that each control device should bear at the current moment based on the actual deviation, thereby realizing dynamic task allocation and collaborative operation among multiple actuators, significantly improving the utilization rate of control resources and reducing the risk of control redundancy and equipment conflict.

[0032] 3. By introducing a disturbance response mechanism, this invention effectively addresses the impact of common non-structural disturbances such as environmental fluctuations, raw material changes, and equipment aging on the stability of the quality control system during production. The quality control system can sense disturbance trends and predict their potential impact on future state evolution, thereby introducing a disturbance adjustment term into the control strategy to achieve early response and dynamic correction to disturbances.

[0033] 4. The control optimization objective function constructed in this invention not only includes the energy cost caused by static deviation, but also the system instability energy caused by disturbance trend. Thus, it unifies state deviation and trend uncertainty into the scope of control optimization, realizes the extension from "control difference" to "control trend", and constructs a higher-level control optimization model. Attached Figure Description

[0034] Figure 1 This is a structural diagram of a quality control system for a pistachio sparkling water production line according to the present invention;

[0035] Figure 2 This is a flowchart of a quality control method for a pistachio sparkling water production line according to the present invention. Detailed Implementation

[0036] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0038] The following description, in conjunction with the accompanying drawings, details a specific solution for a quality control system and method for a pear sparkling water production line provided by the present invention.

[0039] See attached document Figure 1 The diagram illustrates a quality control system structure for a pistachio sparkling water production line according to an embodiment of the present invention. The system includes the following components:

[0040] The system includes a state acquisition module, a state modeling module, a motion gradient calculation module, a disturbance response module, and a control command generation module.

[0041] The status acquisition module periodically acquires various key physical quantities in the pistachio sparkling water production line and transmits the acquired physical quantities to the status modeling module.

[0042] The state modeling module constructs the current state vector of the system based on physical quantities, and constructs the desired state vector based on the target formula parameters stored in the standard process database set within the enterprise. By comparing the current state vector with the desired state vector dimension by dimension, it generates a state deviation vector, and constructs local potential energy functions according to the control weights of the state components defined by the system. The local potential energy functions are summed to construct the total potential energy function. The state vector is then passed to the disturbance response module and the action gradient calculation module; the total potential energy function is then passed to the action gradient calculation module.

[0043] The motion gradient calculation module constructs the motion vector of the control device and calculates the sensitivity of the control device's motion vector to changes in total potential energy through a chain-like differentiation logic structure, generating the motion gradient vector to realize the directional mapping from the state deviation space to the control device's motion space; the control device's motion vector is then passed to the disturbance response module and the control command generation module; the motion gradient vector is then passed to the control command generation module.

[0044] The disturbance response module introduces the influence of external disturbance factors into the control logic of the quality control system to obtain a time-dependent disturbance vector. Based on the second derivative of the state vector and the first derivative of the time-dependent disturbance vector, a disturbance correction vector is constructed. Based on the disturbance correction vector, a disturbance potential energy function is constructed to quantify the potential destructive power of the current disturbance trend on system stability, and the gradient of the effect of the disturbance potential energy function on the control device action vector is calculated to generate a disturbance trend response term. The disturbance trend response term is then passed to the control command generation module.

[0045] The control command generation module merges the action gradient vector with the disturbance trend response term to generate the control gradient vector. Based on the control device action vector and combined with the control gradient vector, the control device action vector is dynamically updated through gradient solving and step size control, which directly acts on the actuator to realize the dynamic closed-loop adjustment of the quality control system state.

[0046] See attached document Figure 2 The diagram illustrates a quality control method for a sparkling water production line according to an embodiment of the present invention, which includes the following steps:

[0047] S1. Collect the physical quantities of the control process, construct the state vector, quantify the deviation of the state vector, construct the total potential energy function, and obtain the total potential energy of the system; construct the action vector of the control equipment, quantify the changing trend of the action vector of the control equipment with respect to the total potential energy of the system, and obtain the action gradient vector.

[0048] In the initial stage of the quality control system operation, the controller communicates with multiple sets of status acquisition ports via a data bus to periodically collect physical quantities directly related to the actual control process, such as the instantaneous flow rate of the pulp pump, the sweetener supply rate, the carbonation injection pressure, the opening degree of the filling valve, and the liquid level in the bottle. These quantities are then uniformly encoded to generate a state vector. ,in, Indicates at time The state vector, Indicates at time The Each dimension of state components It refers to the quantity of physical quantities. Indicates time, Indicates transpose;

[0049] The frequency of physical quantity acquisition is synchronized with the controller's control cycle to ensure that the continuous dynamic modeling conditions within the quality control system meet real-time constraints. To determine whether the current quality control system deviates from a stable operating state, a desired state vector is constructed. This represents the producer's target control value under the current formula control logic, derived from the company's internal standard process database. Indicates the first Target control values ​​for each state dimension;

[0050] To measure the overall deviation between the current state vector and the desired state vector of the quality control system, the state vector and the desired state vector are compared dimension by dimension to obtain the state deviation vector. Based on the state deviation vector, a local potential energy function is constructed, and the local potential energy functions are accumulated to construct the total potential energy function, which is reflected in the controller as an approximate measure of the total amount of actions required by the quality control system. The local potential energy function is defined as follows: ,in, It is the first Each state component at time... The local potential energy value; Indicates the first The weighting coefficients of the influence of each state component on the final product quality, i.e. the control weights of the state components, are determined by statistical regression and have a value range of [0.1, 10]. This is a well-known technique in the art and will not be elaborated here.

[0051] Summing the local potential functions yields the total potential function. Its vector form is ,in It is a state weight coefficient matrix, which contains only the main diagonal elements and represents the control weight of the state component corresponding to each element. It is obtained by statistical regression and the value range is [0.1, 10].

[0052] To derive the regulating load that each control device should bear from the total potential energy function, it is necessary to establish a transmission path between the state vector and the devices. Within the quality control system, the control device action vector is denoted as... ,in, Indicates the first The behavior of a device (such as pump speed, solenoid valve opening, air pressure regulation ratio, etc.) at the current moment The control value, This represents the total number of device behaviors; the influence of control devices on the state vector is modeled as a static linear mapping, expressed as... ,in, It is a device influence mapping matrix, in which the elements represent the linear response coefficients of each device behavior to different state components. It is determined by the experimental perturbation response modeling method, which is a well-known technique to those skilled in the art and will not be described in detail here.

[0053] To obtain the adjustment guidance of each control device's action vector on the changing trend of the current quality control system's total potential energy, it is necessary to calculate the gradient of the total potential energy function with respect to the control device's action vector, thus obtaining the action gradient vector. According to the chain rule, the partial derivative of the total potential energy function with respect to the state vector is first calculated, and then multiplied by the partial derivative of the state vector with respect to the control device's action vector. The specific formula is as follows:

[0054] ,

[0055] in, At any moment The action gradient vector, each component of which can be quantified in the current state, represents the effectiveness of the corresponding device action in reducing the total potential energy. The action gradient vector enables dynamic load distribution and control weight adjustment among control devices, and is the core calculation basis in the parallel adjustment process of multiple actuators.

[0056] S2. Introduce a disturbance response mechanism. Based on the state vector, construct a disturbance correction vector and a disturbance potential energy function. Based on the disturbance potential energy function, construct a control optimization objective function and dynamically update the control device action vector through gradient solving and step size control.

[0057] In a quality control system, while the state vector is affected by the actions of the control equipment, there are also disturbance factors that cannot be directly intervened by the control equipment. These disturbance factors originate from changes in the external environment, fluctuations in the state of raw materials, and abnormal pressures of gas or liquid sources. Physically, they typically manifest as an additional offset to the controlled object and exhibit nonlinear and uncertain characteristics over time. Therefore, a disturbance response mechanism is introduced, representing the disturbance factors as time-dependent disturbance vectors, which correspond one-to-one with the state vectors to represent the instantaneous disturbance impact on each state component.

[0058] The disturbance response mechanism, by introducing a disturbance adjustment term, represents the response feedback quantity generated by the quality control system in response to disturbance trends and state acceleration, i.e., the disturbance correction vector. The disturbance response mechanism is defined as follows:

[0059] ,

[0060] in, At any moment The disturbance adjustment term, i.e. the disturbance correction vector, represents the composite response of the quality control system to the changing trend of the current state vector and the changes of external disturbance factors. It is a state fluctuation sensitivity coefficient matrix. Each element in the state fluctuation sensitivity coefficient matrix represents the amplification factor of the disturbance correction vector caused by the change in acceleration of each state vector. It is constructed by normalizing the standard deviation of the second derivative of the state vector in the historical control cycle. The value range is [0.05, 0.5]. It is a technical means well known to those skilled in the art and will not be described in detail here. It is a disturbance trend gain matrix, which quantifies the degree of intervention of the instantaneous change trend of the time-related disturbance vector on the overall adjustment strategy. The physical meaning of each element in the disturbance trend gain matrix is ​​the equivalent state acceleration generated per unit disturbance rate. It is obtained by fitting the influence of disturbance change on the control effect, and the value range is [0.1,5]. This is a technical means well known to those skilled in the art and will not be elaborated here. Indicates at time The time-dependent perturbation vector. It is the current rate of change of each disturbance source (such as temperature, pressure, feed concentration);

[0061] The technical objective of the entire disturbance regulation term is to provide a quantifiable auxiliary regulation signal that couples the state dimension with the disturbance trend, so that when there is a dynamic disturbance trend in the state vector, the quality control system can generate an early response to the future state evolution direction, thereby achieving a feedforward regulation effect.

[0062] To incorporate the disturbance response into the control optimization objective function, a disturbance potential energy function is constructed to represent the magnitude of the unstable trend energy of the quality control system when driven by a disturbance; the disturbance potential energy function is defined as: This is used to quantify the system fluctuation trend caused by disturbances, reflecting the ability of the current disturbance trend to disrupt the overall regulatory equilibrium state of the quality control system. By combining the disturbance potential energy function and the total potential energy function to form a new control optimization objective function, synchronous control of the overall system state energy and trend energy can be achieved. The specific formula for the final objective function that the quality control system needs to optimize, i.e., the control optimization objective function, is:

[0063] ,

[0064] Control optimization objective function This represents the total potential energy level of the quality control system within the current control cycle, including the impact of static deviations on the energy field and the impact of dynamic disturbances on the evolution trend of the state field. The quality control system must take minimizing the control optimization objective function as the final goal of the control strategy formulation.

[0065] To ensure that the control path is computable and the executed instructions have clear physical operation directions, the quality control system first calculates the control optimization objective function using the chain-reaction method. Regarding the motion vector of control equipment The gradient is used to generate the current control gradient vector; the control gradient vector consists of two parts, namely the system deviation guidance term (i.e., the action gradient vector). With disturbance trend response item These represent the direction of the control response caused by the deviation of the current quality control system state from the optimal target, and the reverse adjustment amount of the control caused by the external disturbance trend on the system state fluctuation, respectively. Together, they constitute the complete logical basis for the quality control system to derive the next adjustment value for equipment operation.

[0066] Furthermore, based on the current control device action vector and combined with the control gradient vector, a control device action vector update formula is constructed to generate the control device action vector for the next control cycle. The specific update formula is as follows:

[0067] ,

[0068] in, It is the control device action vector calculated by the quality control system for the next moment, representing the new control device setpoint that the quality control system will give in the next cycle; The step size coefficient is used to adjust the magnitude of each update of the control device's action vector. The larger the value of the step size coefficient, the faster the control response. It is obtained through online testing and has a value range of [0.01, 0.2]. Indicates at time The control gradient vector; Corresponding to the gradient term in the perturbation trend correction path, the higher-order derivative The third-order trend value, representing the change in system state, is the rate of change of the system parameter acceleration and is used to characterize an early signal of the increasing intensity of oscillations in a mass control system; the second derivative of the disturbance. It represents the acceleration of the disturbance change and is used to reflect the trend of the influence of the disturbance factor on the future direction of change of the quality control system;

[0069] This enables coordinated responses from multiple devices under various state deviations and disturbances, resulting in multi-dimensional parallel control behavior with dynamic adaptive characteristics.

[0070] In summary, a quality control system and method for a pique sparkling water production line have been developed.

[0071] The order of the embodiments is for illustrative purposes only and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0072] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0073] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A quality control method for a sparkling water production line, characterized in that, Includes the following steps: S1. Collect the physical quantities of the control process, construct the state vector, and quantify the degree of deviation of the state vector to obtain the state deviation vector; based on the state deviation vector, construct the local potential energy function; sum the local potential energy functions to construct the total potential energy function and obtain the total potential energy of the system; construct the control equipment action vector, introduce the equipment influence mapping matrix, model the adjustment influence relationship of the control equipment action vector on the state vector as a static linear mapping, and quantify the changing trend of the control equipment action vector on the total potential energy of the system. According to the derivative rule of composite functions, first calculate the partial derivative of the total potential energy function with respect to the state vector, and then multiply it by the partial derivative of the state vector with respect to the control equipment action vector to obtain the action gradient vector. S2. Introduce a disturbance response mechanism: construct a disturbance correction vector based on the state vector; construct a disturbance potential energy function based on the disturbance correction vector; construct a control optimization objective function based on the disturbance potential energy function and the total potential energy function; generate a control gradient vector based on the control optimization objective function; and dynamically update the control device action vector through gradient solving and step size control.

2. The quality control method for a sparkling water production line according to claim 1, characterized in that, S1 specifically includes: Introduce the desired state vector and compare the current state vector with the desired state vector dimension by dimension to obtain the state deviation vector.

3. The quality control method for a sparkling water production line according to claim 1, characterized in that, S2 specifically includes: In the implementation of the disturbance response mechanism, the disturbance factor is represented as a time-dependent disturbance vector. Based on the second derivative of the state vector and the first derivative of the time-dependent disturbance vector, a disturbance correction vector is constructed.

4. The quality control method for a sparkling water production line according to claim 1, characterized in that, S2 specifically includes: The control gradient vector is obtained by calculating the gradient of the control optimization objective function with respect to the control device action vector. The control gradient vector consists of two parts: the action gradient vector and the disturbance trend response term generated based on the disturbance potential energy function.

5. A quality control method for a pistachio sparkling water production line according to claim 4, characterized in that, S2 specifically includes: Based on the current control device action vector and the control gradient vector, the control device action vector for the next moment is generated; the specific formula is as follows: ， in, It is the control device action vector for the next moment; At any moment The control device motion vector; To adjust the step size coefficient; It is the device influence mapping matrix; It is the state weight coefficient matrix; Indicates at time The state vector; It is the desired state vector; It is the state fluctuation sensitivity coefficient matrix; Indicates at time The control gradient vector; It is the perturbation trend gain matrix; Indicates at time The time-dependent perturbation vector.

6. A quality control system for a pique sparkling water production line, applied to the quality control method for a pique sparkling water production line as described in claim 1, characterized in that, Includes the following parts: The system includes a state acquisition module, a state modeling module, a motion gradient calculation module, a disturbance response module, and a control command generation module. The status acquisition module periodically acquires the physical quantities in the pistachio sparkling water production line and transmits the acquired physical quantities to the status modeling module. The state modeling module constructs the current state vector of the system based on physical quantities, and constructs the desired state vector by combining the target formula parameters. By comparing the current state vector with the desired state vector dimension by dimension, it generates a state deviation vector and introduces control weights for the state components to construct a local potential energy function. The local potential energy functions are summed to construct the total potential energy function. The state vector is then passed to the disturbance response module and the action gradient calculation module. Finally, the total potential energy function is passed to the action gradient calculation module. The motion gradient calculation module constructs the motion vector of the control device, calculates the sensitivity of the control device's motion vector to changes in total potential energy, and generates the motion gradient vector; it then transmits the control device's motion vector to the disturbance response module and the control command generation module; finally, it transmits the motion gradient vector to the control command generation module. The disturbance response module introduces the influence of external disturbance factors into the control logic of the quality control system to obtain a time-dependent disturbance vector; based on the second derivative of the state vector and the first derivative of the time-dependent disturbance vector, a disturbance correction vector is constructed; based on the disturbance correction vector, a disturbance potential energy function is constructed, and the gradient of the disturbance potential energy function with respect to the control device action vector is calculated to generate a disturbance trend response term. The disturbance trend response item is passed to the control command generation module; The control command generation module merges the action gradient vector with the disturbance trend response term to generate the control gradient vector; based on the control device action vector at the current moment, and combined with the control gradient vector, the control device action vector is dynamically updated through gradient solving and step size control.

Images