Industrial computer control method based on internet of things

Abstract

The application relates to the technical field of intelligent control, and discloses an industrial computer control method based on an Internet of Things, which comprises the following steps: first, an industrial computer dynamic characteristic state space model is established, the running state, control input and external disturbance data of the industrial computer are obtained through measurement, and the control input delay and parameter uncertainty are determined by using a clock and a network monitoring tool; second, offline simulation is carried out by using the state space model and parameter data obtained in the first step, a delay prediction model and a robust feedback control law are obtained by feedback control gain design, and it is verified that the closed-loop system meets the stability requirement in the offline simulation environment. The whole-process closed-loop control technical scheme is adopted, the real-time monitoring of the industrial computer state and the delay adaptive compensation effect are realized, and compared with the scheme in the prior art that data acquisition lags behind and control response is not timely, the problems of system response delay and insufficient stability are solved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and in particular to an industrial computer control method based on the Internet of Things. Background Art

[0002] In traditional industrial control systems, industrial computer control methods face technical challenges, particularly in data acquisition and delay compensation. Traditional control methods often rely on offline models and manual adjustments, resulting in long system response times and an inability to handle changing control inputs and external disturbances in real time. Furthermore, due to network instability, control signals are often subject to delays, severely impacting the real-time performance and stability of industrial computers. Consequently, existing industrial computer control methods suffer from the following deficiencies:

[0003] Existing technologies suffer from data acquisition and control response lags, making them unable to adapt to rapidly changing industrial control needs. Traditional industrial computer systems rely primarily on scheduled data acquisition and offline calculations, resulting in control systems unable to respond promptly to changing conditions in real-time. Due to the influence of control input and delays, the system cannot make precise adjustments, resulting in response lag and poor stability.

[0004] Among existing technical solutions, many industrial computer control methods rely on fixed control algorithms and parameters, lacking dynamic adaptive capabilities. Traditional methods employ fixed control strategies, making it difficult for the system to effectively adjust to complex and unexpected operating conditions. They also neglect the integration of delay compensation and dynamic regulation, resulting in weak interference immunity in practical applications.

[0005] Most current industrial computer control systems lack efficient online optimization mechanisms. Traditional control strategies typically rely on offline simulation and preset parameters for control adjustments, failing to dynamically optimize the system based on real-time data. In complex environments, control strategies can fail due to input signal fluctuations and external interference, leading to system instability and even the risk of system crashes.

[0006] Existing methods fail to effectively integrate delay compensation and robust control, resulting in reduced accuracy in the system's response to delays. While traditional approaches consider delay compensation, they focus solely on a single control strategy and lack robustness, leading to reduced stability in the IPC's response and even impacting the overall performance of the ICS.

[0007] The above analysis shows that existing technologies fail to meet the requirements of modern industrial control systems for high real-time performance, dynamic adjustment and robustness in many aspects.

[0008] Therefore, those skilled in the art provide an industrial computer control method based on the Internet of Things to solve the above-mentioned problems. Summary of the Invention

[0009] In view of the deficiencies in the prior art, the present invention provides an industrial computer control method based on the Internet of Things to solve the problems raised in the above background technology.

[0010] To achieve the above objectives, the present invention is implemented through the following technical solutions: A method for controlling an industrial computer based on the Internet of Things, comprising the following steps:

[0011] The first step is to build a state-space model of the IPC's dynamic characteristics. This involves measuring the IPC's operating status, control inputs, and external disturbance data. Clock and network monitoring tools are then used to determine control input delays and parameter uncertainties.

[0012] In the second step, offline simulation is performed using the state-space model and parameter data obtained in the first step. The delay prediction model and robust feedback control law are obtained through feedback control gain design, and the closed-loop system is verified to meet stability requirements in the offline simulation environment.

[0013] In the third step, based on the feedback control law and delay prediction model obtained in the second step, the control input is dynamically solved using an optimization method within the online model predictive control framework.

[0014] The fourth step is to collect the actual operating status, control input, and control delay data of the industrial computer based on the control input solved in the third step, and to monitor and dynamically update the network transmission delay in real time;

[0015] In the fifth step, the online model predictive control algorithm obtained through real-time data collection and monitoring in the fourth step is embedded into the industrial computer and edge computing platform to achieve the integration of state collection, delay monitoring and online optimization control modules;

[0016] In the sixth step, based on the control method deployed in the fifth step, field tests were conducted to verify that the IPC’s response, overshoot, and steady-state error met the design requirements. A self-diagnostic safety monitoring module was established to automatically trigger the safety control mode when abnormal delays and external disturbances were detected, enabling the IPC to maintain stable operation under abnormal conditions.

[0017] Preferably, in the step of establishing the state space model, the dynamic characteristics of the industrial computer are realized by a mathematical model that describes the control input, state vector and external disturbance vector. The model combines the control signal transmission delay and parameter uncertainty, and adopts a local linearization method near the system operating point to reflect the dynamic response of the industrial computer.

[0018] Preferably, in the offline simulation and feedback control gain design step, the delay prediction model uses exponential matrix transformation and integral operation to describe the effect of delay compensation, and the model is implemented by the following formula:

[0019]

[0020] Among them, A D is the system dynamic matrix, B E is the input matrix, x A (t) is the system status,

[0021] is the control input, τ C is the input delay, θ O is the prediction time step, x A (t+θ O ) represents the predicted time t+θ O The state vector of the industrial computer is s, which represents the integral variable.

[0022] Indicates that at time t+s-τ C The control input vector actually applied to the industrial computer after delay compensation is shown below, where e is the exponent.

[0023] Preferably, the robust feedback control law design step obtains the state feedback gain and delay compensation gain of the system by solving the Riccat i inequality, and the solution formula of the Riccat i inequality is as follows:

[0024]

[0025] Among them, P R is a positive definite matrix, is the state feedback matrix, R L is the input weight matrix, Q K is the state penalty matrix, γ S is the disturbance coefficient, is the identity matrix,

[0026] Represents the industrial computer dynamic matrix A D The transpose of B E represents the input matrix, Represents the input weight matrix R L The inverse matrix of .

[0027] Preferably, in the feedback control law design, the delay compensation gain is calculated by the following formula:

[0028]

[0029] Among them, L Q (τ C ) is the delay compensation matrix, τ C To control input delay, is the state feedback matrix, Represents the input weight matrix R L The inverse matrix of Represents the input matrix B E The transpose of P R represents a positive definite matrix, Represents the matrix exponential operation, A D represents the dynamic matrix of the industrial computer, τ C Indicates control input delay.

[0030] Preferably, in the online model predictive control framework, the optimization performance index function includes the following parts:

[0031]

[0032] Where J represents the performance index function, t0 represents the start time of control optimization, t f represents the control optimization termination time, s represents the integral variable, x A (s) represents the state vector of the industrial computer at time s, reflecting the dynamic response of the industrial computer.

[0033] u B (s) represents the control input vector applied by the industrial computer at time s to reflect the external input effect, α M represents the delay penalty coefficient, τ C To control input delay, R L is the input weight matrix, Q K is the state penalty matrix.

[0034] Preferably, in the online model predictive control step, the optimization method dynamically solves the optimal control input according to the current state of the system and the delay information, and the optimal control input is calculated by the following formula:

[0035]

[0036] in, It means that the optimal control input vector obtained by online optimization method at time t reflects the actual control signal applied by the industrial computer.

[0037] K P is the state feedback matrix, L Q (τ C ) is the delay compensation matrix, τ C To control input delay, x A (t represents the state vector of the industrial computer at the prediction time t,

[0038] x A )t-τ C ) represents the predicted time t-τ C The state vector of the industrial computer.

[0039] Preferably, in the real-time data collection and delay monitoring steps, by deploying sensors and network monitoring modules on the industrial computer, the system status information, control input and delay data are monitored in real time, and the parameters of the delay prediction model are dynamically adjusted through the edge computing platform to enable the online controller to adapt to changes in network delay.

[0040] Preferably, in the step of edge node deployment and field testing, the control algorithm is embedded and tested in the following manner:

[0041] The control algorithm verified by offline simulation is transplanted to the industrial computer and edge computing platform, and long-term experimental data recording and testing are carried out to verify the response time, overshoot, steady-state error and stability of the control method in the actual industrial control environment.

[0042] Preferably, in the safety monitoring and exception handling steps, a self-diagnosis module is integrated for real-time monitoring. If it is detected that the control input delay and external disturbance exceed the preset threshold, the safety control mode and alarm mechanism are automatically triggered to ensure that the industrial control system can maintain stable operation under abnormal conditions and prevent the expansion of system failures.

[0043] The present invention provides an industrial computer control method based on the Internet of Things. It has the following beneficial effects:

[0044] 1. The present invention adopts a full-process closed-loop control technology solution to achieve real-time monitoring of the industrial computer status and delay adaptive compensation effect. Compared with the existing technology with data acquisition lag and untimely control response, it solves the problems of system response delay and insufficient stability.

[0045] 2. The present invention adopts a data acquisition and edge deployment technology solution based on the Internet of Things to achieve real-time optimization and dynamic adjustment of control input. Compared with the offline controller solution in the existing technology with slow update speed and poor adaptability, it solves the shortcomings of poor real-time performance and dynamic response.

[0046] 3. The present invention adopts a technical solution that integrates robust feedback control and delay compensation to achieve dual correction effects of state feedback and delay compensation. Compared with the existing solution that focuses on a single control strategy, it solves the problems of unbalanced dynamic response and insufficient anti-interference ability of industrial computers.

[0047] 4. The present invention adopts an online model predictive control algorithm technology solution to achieve dynamic solution of optimal control input and real-time optimization effect. Compared with the solution of fixed control strategy prone to failure in the existing technology, it solves the shortcomings of control adjustment lag and poor system stability under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 Flowchart of the present invention. DETAILED DESCRIPTION

[0049] To help those skilled in the art understand the present invention, the following will provide a clear and complete description of the technical solutions in the embodiments of the present invention, in conjunction with the accompanying drawings. Obviously, the described embodiments are only partial embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0050] The present invention is described in detail below with reference to the accompanying drawings:

[0051] Example:

[0052] Please see the attached Figure 1 The embodiment of the present invention provides an industrial computer control method based on the Internet of Things, comprising the following steps:

[0053] The first step is to establish a state-space model of the IPC's dynamic characteristics. This involves measuring the IPC's operating status, control inputs, and external disturbance data, and using clock and network monitoring tools to determine control input delays and parameter uncertainties. During this state-space modeling step, the IPC's dynamic characteristics are captured using a mathematical model that describes the control inputs, state vectors, and external disturbance vectors. This model incorporates control signal transmission delays and parameter uncertainties, and employs a local linearization approach near the system's operating point to reflect the IPC's dynamic response.

[0054] In the second step, offline simulation is performed using the state-space model and parameter data obtained in the first step. The delay prediction model and robust feedback control law are obtained through feedback control gain design, and the closed-loop system is verified to meet stability requirements in the offline simulation environment.

[0055] In the third step, based on the feedback control law and delay prediction model obtained in the second step, the control input is dynamically solved using an optimization method within the online model predictive control framework.

[0056] The fourth step is to collect the actual operating status, control input, and control delay data of the industrial computer based on the control input solved in the third step, and to monitor and dynamically update the network transmission delay in real time;

[0057] In the fifth step, the online model predictive control algorithm obtained by real-time data acquisition and monitoring in the fourth step is embedded into the industrial computer and the edge computing platform to realize the integration of state acquisition, delay monitoring and online optimization control module. In the real-time data acquisition and delay monitoring step, the state information, control input and delay data of the system are monitored in real time by deploying sensors and network monitoring modules on the industrial computer, and the parameters of the delay prediction model are dynamically adjusted through the edge computing platform to enable the online controller to adapt to changes in network delay. In the edge node deployment and field test step, the embedding and testing of the control algorithm are carried out in the following way:

[0058] The control algorithm verified by offline simulation is transplanted to the industrial computer and the edge computing platform for long-term experimental data recording and testing to verify the response time, overshoot, steady-state error and stability of the control method in the actual industrial control environment.

[0059] In the sixth step, according to the control method deployed in the fifth step, the response of the industrial computer, overshoot and steady-state error are verified through field testing to meet the design requirements, and a self-diagnosis safety monitoring module is set up to automatically trigger a safety control mode when abnormal delay and external disturbance are detected to maintain stable operation of the industrial computer under abnormal conditions. In the safety monitoring and abnormal handling step, the self-diagnosis module is integrated for real-time monitoring. If the control input delay and external disturbance exceed the preset threshold, the safety control mode and alarm mechanism are automatically triggered to ensure that the industrial control system can maintain stable operation under abnormal conditions and prevent system failure from expanding.

[0060] The embodiment adopts a technical solution combining real-time state acquisition of the industrial computer, offline simulation and online model predictive control, uses the edge computing platform to process the collected data, dynamically calculates the optimal control input, and monitors the delay and disturbance through the self-diagnosis safety module. This solution breaks through the difficulties of traditional offline parameter setting, control response lag and insufficient delay compensation, and realizes real-time dynamic control and automatic compensation.

[0061] In the implementation process, first, the industrial computer is comprehensively parameterized. Clock, sensor and network monitoring tools are deployed to collect the operating state, control input and external disturbance data of the industrial computer, and to detect signal transmission delay and parameter uncertainty. Through data analysis, a state space model describing the relationship between control input, state vector and disturbance vector is established. This model uses a local linearization method to ensure high accuracy in the vicinity of the operating point, laying a foundation for subsequent steps.

[0062] Based on the established state-space model, the system was fully simulated using offline simulation. During this process, a delay prediction model and robust feedback control law were derived through feedback control gain design. Delay compensation was achieved using exponential matrix transformation and integral operations, and the positive definite matrix, state feedback gain, and delay compensation gain were obtained by solving the Riccati inequality. This process fully accounted for the effects of control signal transmission delays and external disturbances on the IPC state, ensuring that the stability and responsiveness of the closed-loop system met the design requirements during the simulation.

[0063] After successful offline simulation verification, the feedback control law and delay prediction model were introduced into the online model predictive control framework. Optimization calculations were performed using the edge computing platform to solve for the optimal control input in real time. This solution adjusts the optimal input based on the currently collected state data and detected network delay information, achieving dynamic compensation of the control signal. The controller uses an optimized performance indicator function to balance state error, control input, and delay penalty through integration to achieve global optimal control.

[0064] During online control, the system collects the actual operating status, control inputs, and delay data from the industrial computer, and monitors delays in real time through a deployed network monitoring module. The edge platform dynamically updates the delay prediction model parameters based on the latest data, ensuring that the online controller always uses accurate data for calculations. The control algorithm is embedded in the industrial computer and the edge computing platform, forming an integrated module that combines status collection, delay monitoring, and online optimal control. This ensures that the control process maintains real-time responsiveness and stability even in complex network environments.

[0065] To ensure safe operation of the industrial control system under abnormal circumstances, this embodiment incorporates a self-diagnostic safety monitoring module. This module monitors control input delays and external disturbances in real time. Upon detecting an abnormality exceeding a preset threshold, it immediately triggers a safety control mode and an alarm mechanism. The safety module employs simple and straightforward logic to ensure that the system can quickly switch to a safe state in an emergency, preventing the spread of faults and maintaining overall system security and stability.

[0066] This embodiment combines real-time state acquisition with offline simulation and employs online model predictive control, significantly improving the response speed of the industrial control computer. It utilizes the edge platform to dynamically optimize control inputs and implement adaptive delay compensation, ensuring stable system operation despite network delays. An integrated self-diagnostic safety monitoring module provides automatic protection against abnormal conditions, enhancing system robustness. The overall solution is compact and simple to implement, significantly reducing the data acquisition lag and untimely control response issues of traditional industrial control systems.

[0067] In the offline simulation and feedback control gain design steps, the delay prediction model uses exponential matrix transformation and integral operations to describe the effect of delay compensation. The model is implemented using the following formula:

[0068]

[0069] Among them, A D is the system dynamic matrix, B E is the input matrix, x A (t) is the system status,

[0070] is the control input, τ C is the input delay, θ O is the prediction time step, x A (t+θ O ) represents the predicted time t+θ O The state vector of the industrial computer is s, which represents the integral variable.

[0071] Indicates that at time t+s-τ C The control input vector actually applied to the industrial computer after delay compensation is shown below, where e is the exponent.

[0072] Through exponential matrix transformation and integral operation, effective compensation for control input delay is achieved, so that the system can maintain stable operation under different delay environments and reduce oscillation and overshoot.

[0073] By introducing the delay prediction model, the system can accurately calculate the optimal control input, ensure that the change of the state vector conforms to the set trajectory, and improve the overall control accuracy.

[0074] The prediction model can dynamically adjust the compensation parameters over time, allowing the controller to adapt to different working conditions and delay changes, thereby improving the robustness and adaptability of the system.

[0075] The mathematical method of exponential matrix and integral operation is adopted to make the delay compensation calculation process efficient, avoid the complex numerical solution process, and improve the real-time performance of the control algorithm.

[0076] In summary, through exponential matrix transformation and integral operations, we achieve precise compensation for control input delay, effectively improving the real-time responsiveness and stability of industrial computers. Compared with traditional methods, this approach reduces control computational complexity and improves dynamic adaptability, enabling industrial control systems to maintain efficient and stable operation in complex environments.

[0077] The robust feedback control law design step obtains the system's state feedback gain and delay compensation gain by solving the Riccat i inequality. The solution formula for the Riccat i inequality is as follows:

[0078]

[0079] Among them, P R is a positive definite matrix, is the state feedback matrix, R L is the input weight matrix, Q K is the state penalty matrix, γ S is the disturbance coefficient, is the identity matrix,

[0080] Represents the industrial computer dynamic matrix A D The transpose of B E represents the input matrix, Represents the input weight matrix R L The inverse matrix of .

[0081] By solving the Riccat i inequality, the optimal state feedback gain can be obtained, which can keep the system stable under the influence of external disturbances and delays and avoid system divergence and unstable oscillation.

[0082] The solution to the Riccat i inequality provides the optimal state penalty and input weight distribution, ensuring that the system can achieve accurate state regulation under limited control input conditions, reducing overshoot and steady-state error.

[0083] The calculation of state feedback gain and delay compensation gain can optimize the control input while ensuring system performance, minimize energy consumption and improve the system's operating efficiency.

[0084] In the presence of uncertain disturbances and model errors, the control input can be automatically adjusted through the robust feedback control law, so that the system has good anti-interference ability in complex environments.

[0085] In summary, this invention solves the Riccati inequality to design optimal state feedback gains and delay compensation gains, achieving precise control and enhanced robustness for industrial computers. Compared to traditional control methods, this solution maintains system stability under various operating conditions, optimizes energy consumption, and improves interference rejection, making industrial computers more adaptable and reliable in complex industrial environments.

[0086] In the feedback control law design, the delay compensation gain is calculated by the following formula:

[0087]

[0088] Among them, L Q (τ C ) is the delay compensation matrix, τ C To control input delay, is the state feedback matrix, Represents the input weight matrix R L The inverse matrix of Represents the input matrix B E The transpose of PR represents a positive definite matrix, Represents the matrix exponential operation, A D represents the dynamic matrix of the industrial computer, τ C Indicates control input delay.

[0089] By using matrix exponential operations, the delay compensation gain can be accurately calculated, so that the control input can effectively offset the impact caused by network delay and improve control accuracy.

[0090] This method can dynamically adjust the compensation gain so that the system can adapt to different degrees of delay changes, optimize the dynamic response of the industrial computer, and avoid control errors caused by hysteresis.

[0091] An optimization method combining the state feedback matrix with the input weight matrix is ​​adopted to make the control system stable under the influence of external disturbances and delay uncertainties, and improve the anti-disturbance capability.

[0092] Through the mathematical optimization method of matrix exponential operation, the complex iterative solution process is avoided, the computational complexity is reduced, and the real-time performance and substantiality of the control algorithm are improved.

[0093] In summary, calculating the delay compensation gain through matrix exponential operations allows for precise compensation of network delays, improving the system's dynamic responsiveness and stability. Compared to traditional fixed compensation strategies, this method dynamically adjusts compensation parameters based on actual delay conditions, enabling industrial computers to maintain efficient and stable operation in complex network environments.

[0094] In the online model predictive control framework, the optimization performance indicator function includes the following parts:

[0095]

[0096] Where J represents the performance index function, t0 represents the start time of control optimization, t f represents the control optimization termination time, s represents the integral variable, x A (s) represents the state vector of the industrial computer at time s, reflecting the dynamic response of the industrial computer.

[0097] u B (s) represents the control input vector applied by the industrial computer at time s to reflect the external input effect, α M represents the delay penalty coefficient, τ C To control input delay, R L is the input weight matrix, Q K is the state penalty matrix.

[0098] The optimized performance indicator function ensures that the control strategy of the industrial computer reduces unnecessary energy consumption while meeting the target performance by balancing state error, control input and delay penalty.

[0099] The control process is globally optimized by the integral method, so that the control system remains stable in long-term operation, overshoot and steady-state error are reduced, and overall stability is improved.

[0100] This method can dynamically adjust the control optimization target according to the real-time collected data, so that the industrial control computer can adapt to different working conditions and external disturbances and improve robustness.

[0101] The optimization performance index function adopts mathematical modeling to improve the calculation efficiency, make the optimal control input solution fast and accurate, and ensure the real-time performance of the system.

[0102] In summary, by optimizing the performance indicator function and comprehensively considering state error, control input, and delay penalty, we achieve precise control and dynamic optimization of industrial computers. Compared with traditional control methods, this solution can maintain stable operation under different operating conditions and complex network environments, improve response speed, and optimize system energy consumption, making industrial control systems more intelligent and flexible.

[0103] In the online model predictive control step, the optimization method dynamically solves the optimal control input based on the current state of the system and the delay information. The optimal control input is calculated using the following formula:

[0104]

[0105] in, It means that the optimal control input vector obtained by online optimization method at time t reflects the actual control signal applied by the industrial computer.

[0106] K P is the state feedback matrix, L Q (τ C ) is the delay compensation matrix, τ C To control input delay, x A (t represents the state vector of the industrial computer at the prediction time t,

[0107] x A (t-τ C ) represents the predicted time t-τ C The state vector of the industrial computer.

[0108] The optimization method calculates the optimal control input in real time based on the current state of the system and network latency, ensuring that the control strategy always meets the best performance and improving the response speed and accuracy of the system.

[0109] The delay compensation matrix is ​​used to enable the control input to be dynamically adjusted according to the actual network delay, reducing the impact of delay on system stability and enhancing the robustness of the system.

[0110] The method can analyze and predict the state information in real time, so that the industrial computer can adapt to different working conditions and environmental changes, and ensure the control precision and system stability.

[0111] The optimal solution ensures that the control input can meet the system performance requirements and reduce unnecessary energy consumption, thereby improving the overall efficiency and economy of the control system.

[0112] In summary, by using the online optimization method, the optimal control input is dynamically calculated based on the current state and delay information of the system, thereby realizing precise control and intelligent adjustment of the industrial computer. Compared with the traditional fixed control strategy, this method can adaptively adjust according to real-time data, optimize control precision, improve system response speed, and enhance anti-interference ability, so that the industrial control system can run efficiently and stably in complex environments.

[0113] Although embodiments of the present application have been shown and described, it will be understood by those having ordinary skill in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the present application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for controlling an industrial computer based on the Internet of Things, characterized in that: The steps include: The first step is to build a state-space model of the IPC's dynamic characteristics. This involves measuring the IPC's operating status, control inputs, and external disturbance data. Clock and network monitoring tools are then used to determine control input delays and parameter uncertainties. In the second step, offline simulation is performed using the state-space model and parameter data obtained in the first step. The delay prediction model and robust feedback control law are obtained through feedback control gain design, and the closed-loop system is verified to meet stability requirements in the offline simulation environment. The model is implemented by the following formula: Among them, A D is the system dynamic matrix, B E is the input matrix, x A (t) is the system status, is the control input, τ C is the input delay, θ O is the prediction time step, x A (t+θ O ) represents the predicted time t+θ O The state vector of the industrial computer is s, which represents the integral variable. Indicates that at time t+s-τ C The control input vector actually applied to the industrial computer after delay compensation is shown below, where e is the exponent; In the third step, based on the feedback control law and delay prediction model obtained in the second step, the control input is dynamically solved using an optimization method within the online model predictive control framework. Among them, the functions involved in optimizing performance indicators include the following parts: Where J represents the performance index function, t0 represents the start time of control optimization, t f represents the control optimization termination time, s represents the integral variable, x A (s) represents the state vector of the industrial computer at time s, reflecting the dynamic response of the industrial computer. u B (s) represents the control input vector applied by the industrial computer at time s to reflect the external input effect, α M represents the delay penalty coefficient, τ C To control input delay, R L is the input weight matrix, Q K is the state penalty matrix; The fourth step is to collect the actual operating status, control input, and control delay data of the industrial computer based on the control input solved in the third step, and to monitor and dynamically update the network transmission delay in real time; In the fifth step, the online model predictive control algorithm obtained through real-time data collection and monitoring in the fourth step is embedded into the industrial computer and edge computing platform to achieve the integration of state collection, delay monitoring and online optimization control modules; In the sixth step, based on the control method deployed in the fifth step, field tests were conducted to verify that the IPC’s response, overshoot, and steady-state error met the design requirements. A self-diagnostic safety monitoring module was established to automatically trigger the safety control mode when abnormal delays and external disturbances were detected, enabling the IPC to maintain stable operation under abnormal conditions.

2. The method for controlling an industrial computer based on the Internet of Things according to claim 1, characterized in that: In the step of establishing the state space model, the dynamic characteristics of the industrial computer are realized by describing a mathematical model of control input, state vector and external disturbance vector. The model combines control signal transmission delay and parameter uncertainty, and adopts a local linearization method near the system operating point to reflect the dynamic response of the industrial computer.

3. The method for controlling an industrial computer based on the Internet of Things according to claim 1, characterized in that: The robust feedback control law design step obtains the state feedback gain and delay compensation gain of the system by solving the Riccati inequality. The solution formula of the Riccati inequality is as follows: Among them, P R is a positive definite matrix, is the state feedback matrix, R L is the input weight matrix, Q K is the state penalty matrix, γ S is the disturbance coefficient, is the identity matrix, Represents the industrial computer dynamic matrix A D The transpose of B E represents the input matrix, Represents the input weight matrix R L The inverse matrix of .

4. The method for controlling an industrial computer based on the Internet of Things according to claim 3, characterized in that: In the feedback control law design, the delay compensation gain is calculated by the following formula: Among them, L Q (τ C ) is the delay compensation matrix, τ C To control input delay, is the state feedback matrix, Represents the input weight matrix R L The inverse matrix of Represents the input matrix B E The transpose of P R represents a positive definite matrix, Represents the matrix exponential operation, A D represents the dynamic matrix of the industrial computer, τ C Indicates control input delay.

5. The method for controlling an industrial computer based on the Internet of Things according to claim 1, characterized in that: In the online model predictive control step, the optimization method dynamically solves the optimal control input based on the current state of the system and delay information. The optimal control input is calculated using the following formula: in, It means that the optimal control input vector obtained by online optimization method at time t reflects the actual control signal applied by the industrial computer. K P is the state feedback matrix, L Q (τ C ) is the delay compensation matrix, τ C To control input delay, x A (t) represents the state vector of the industrial computer at the predicted time t, x A (t-τ C ) represents the predicted time t-τ C The state vector of the industrial computer.

6. The method for controlling an industrial computer based on the Internet of Things according to claim 1, characterized in that: In the real-time data collection and delay monitoring step, sensors and network monitoring modules are deployed on the industrial computer to monitor the system status information, control input and delay data in real time, and the parameters of the delay prediction model are dynamically adjusted through the edge computing platform to enable the online controller to adapt to changes in network delay.

7. The method for controlling an industrial computer based on the Internet of Things according to claim 1, characterized in that: The algorithm embedding in the fifth step and the field test in the sixth step are achieved by the following methods: The control algorithm verified by offline simulation is transplanted to the industrial computer and edge computing platform, and long-term experimental data recording and testing are carried out to verify the response time, overshoot, steady-state error and stability of the control method in the actual industrial control environment.

8. The method for controlling an industrial computer based on the Internet of Things according to claim 1, characterized in that: In the safety monitoring and exception handling steps, an integrated self-diagnosis module performs real-time monitoring. If it is detected that the control input delay and external disturbance exceed the preset threshold, the safety control mode and alarm mechanism are automatically triggered to ensure that the industrial control system can maintain stable operation under abnormal conditions and prevent the expansion of system failures.

Images