Servo hydraulic system optimization control method and system for press test platform

Abstract

The present invention provides a servo hydraulic system optimization control method and system for a press test platform, which relates to the technical field of presses, including: based on dynamic characteristics, by deriving mathematical expressions of various components of the system, constructing a state space expression, and establishing a mathematical model of the servo hydraulic system; constructing a predictive control algorithm based on the mathematical model, by establishing a state space prediction model and determining a multi-objective optimization function, using a rolling optimization strategy to solve the quadratic programming problem, obtaining an optimal control sequence, and using the first element of the optimal control sequence as the control input at the current moment; introducing feedforward compensation and pressure feedback compensation, wherein the feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, and the pressure feedback compensation designs a digital filter based on the pressure error, and the feedforward compensation signal is merged with the pressure feedback compensation signal, and the signal is superimposed with the control signal to obtain a final control signal.

Description

Technical Field

[0001] The present invention relates to the technical field of presses, and in particular to a servo hydraulic system optimization control method and system for a press test platform. Background Art

[0002] A press is a device that uses the hydraulic transmission principle to apply pressure to a workpiece through the pressure of hydraulic oil. It is widely used in the fields of material forming, forging, and assembly. The control performance of a press directly affects product quality and production efficiency, so higher and higher requirements are placed on the press control system.

[0003] Traditional press control systems mainly use PID control algorithms, and manually adjust PID parameters to control process variables such as pressure and displacement. However, the press system has serious nonlinearity, time-varying and uncertainty. It is difficult for traditional PID control to achieve satisfactory control effects. The system has slow dynamic response speed, low steady-state accuracy and poor robustness. However, advanced control strategies based on intelligent control theory still have some shortcomings. The establishment of fuzzy control rules relies on expert experience. It is difficult to achieve adaptive adjustment in the face of unknown working conditions. The adaptive control algorithm is complex, the real-time calculation amount is large, and it is not easy to implement in engineering.

[0004] In summary, there is an urgent need for a new type of press servo hydraulic system optimization control method that can comprehensively apply advanced control theory and intelligent optimization algorithm to achieve high precision, high efficiency and high reliability of the press control system. The present invention can solve the problems in the prior art. Summary of the invention

[0005] The embodiment of the present invention provides a servo hydraulic system optimization control method and system for a press test platform, which can solve the problems in the prior art.

[0006] According to a first aspect of the embodiments of the present invention,

[0007] Provided is a servo hydraulic system optimization control method for a press test platform, comprising:

[0008] Based on the dynamic characteristics of the servo valve, hydraulic cylinder and load, the mathematical model of the servo hydraulic system is established by deriving the mathematical expressions of each component of the system and constructing a state space expression containing multiple state variables;

[0009] Based on the mathematical model, a predictive control algorithm is constructed, by establishing a state space prediction model and determining a multi-objective optimization function, a rolling optimization strategy is adopted to solve the quadratic programming problem, an optimal control sequence is obtained, and the first element of the optimal control sequence is used as the control input at the current moment;

[0010] Feedforward compensation and pressure feedback compensation are introduced into the predictive control algorithm. The feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal. The pressure feedback compensation designs a digital filter based on the pressure error, merges the feedforward compensation signal with the pressure feedback compensation signal, and superimposes them with the control signal to obtain the final control signal.

[0011] In an optional embodiment,

[0012] Based on the dynamic characteristics of the servo valve, hydraulic cylinder and load, the mathematical expressions of the system components are derived, and the state space expression containing multiple state variables is constructed. The mathematical model of the servo hydraulic system is established, including:

[0013] Based on the structure and working process of the servo valve, the corresponding relationship between the valve core displacement and the flow rate of the servo valve is established, the flow equation is determined, the relationship between the valve core displacement and the control signal is constructed using a second-order vibration link model, the dynamic characteristic equation is determined, and the servo valve mathematical model is established by combining the flow equation and the dynamic characteristic equation;

[0014] Based on the relationship between the temperature change of the hydraulic oil and the bulk elastic modulus, a continuity equation is determined, the friction force is determined using the Stribeck friction model, a force balance equation is constructed, and a mathematical model of the hydraulic cylinder is established by combining the continuity equation and the force balance equation;

[0015] Based on the nonlinear characteristics of the load, nonlinear terms are generated, motion equations are constructed, and load mathematical models are established;

[0016] Determine state variables based on the servo valve mathematical model, the hydraulic cylinder mathematical model and the load mathematical model, wherein the state variables include valve core displacement, valve core speed, piston displacement, piston speed, rodless chamber pressure, rod chamber pressure, load displacement and load speed;

[0017] Establish the relationship between the time derivative of the state variable and the input quantity, generate the state equation, establish the relationship between the output quantity and the state variable and the input quantity, generate the output equation, construct the state space expression based on the state equation and the output equation, and determine the mathematical model of the servo hydraulic system accordingly.

[0018] In an optional embodiment,

[0019] The servo valve mathematical model includes:

[0020] The flow equation is as follows:

[0021] ;

[0022] in, Q Indicates the servo valve flow rate,K q represents the flow gain coefficient, x v Indicates the displacement of the servo valve core, Δ p Indicates the pressure difference before and after the servo valve;

[0023] The dynamic characteristic equation is as follows:

[0024] ;

[0025] in, X v ( s ) represents the Laplace transform of the servo valve spool displacement, I ( s ) represents the Laplace transform of the control signal, K i represents the current displacement gain coefficient, ω n represents the natural frequency of the servo valve, ζ represents the damping ratio, s represents a complex variable in the Laplace transform.

[0026] In an optional embodiment,

[0027] The hydraulic cylinder mathematical model includes:

[0028] The continuity equation is as follows:

[0029] ;

[0030] in, Q 1 indicates the flow rate of the rodless cavity, Q 2 indicates the flow rate of the rod cavity. A 1 represents the effective area of ​​the rodless cavity, A 2 represents the effective area of ​​the rod cavity. x p represents the piston displacement, t Indicates time, V 1 represents the initial volume of the rodless cavity, V 2 represents the initial volume of the rod cavity, β e represents the effective bulk modulus of the liquid, p 1 indicates the rodless cavity pressure, p 2 indicates rod cavity pressure, C ip represents the internal leakage coefficient;

[0031] The force balance equation is as follows:

[0032] ;

[0033] in, m represents the equivalent mass of the piston and the load, B p represents the equivalent damping coefficient of the piston and the load, K s represents the equivalent stiffness of the hydraulic cylinder, F L Represents external load force.

[0034] In an optional embodiment,

[0035] The load mathematical model includes:

[0036] The equation of motion is as follows:

[0037] ;

[0038] in, m L Indicates the load mass, x L represents the load displacement, B L represents the load damping factor, K L represents the load stiffness, F P Indicates the output force of the hydraulic cylinder.

[0039] In an optional embodiment,

[0040] Based on the mathematical model, a predictive control algorithm is constructed. By establishing a state space prediction model and determining a multi-objective optimization function, a rolling optimization strategy is used to solve the quadratic programming problem, and the optimal control sequence is obtained. The first element of the optimal control sequence is used as the control input at the current moment, including:

[0041] Based on the mathematical model, a continuous-time state-space expression is determined, and a discrete-time state-space model is obtained after discretization.

[0042] Based on the discrete time state space model, according to the preset prediction time domain, the state of multiple future time steps is calculated, the output prediction equation is derived, the prediction result is obtained, and the prediction result is decomposed into a free response and a forced response;

[0043] According to the preset control target, based on the free response and the forced response, a multi-objective optimization function including a state error index, a control energy index and a control increment index is generated; based on the predetermined physical characteristics and the preset control requirements, based on the discrete time state space model and the output prediction equation, constraint conditions including control input constraints, control increment constraints, state variable constraints and output variable constraints are determined;

[0044] Expand the multi-objective optimization function into a quadratic form, generate state error terms, control input terms and control increment terms, and construct a Hessian matrix and a gradient vector in combination with the state control cross terms that determine the coupling relationship, and determine the objective function of the quadratic programming problem; based on the constraint conditions, determine linear inequality constraints and linear equality constraints, and generate a constraint matrix and a constraint vector in combination with the state prediction matrix and the control matrix generated based on the discrete time state space model and the prediction time domain, and standardize the quadratic programming problem;

[0045] Based on the preprocessing of the Hessian matrix through the Cholesky decomposition method, the optimal solution of the previous time step is rolled to the current time step as the starting solution, a new state is obtained, the discrete-time state space model is updated, and the standardized quadratic programming problem is iteratively solved to obtain the optimal control sequence.

[0046] In an optional embodiment,

[0047] Feedforward compensation and pressure feedback compensation are introduced into the predictive control algorithm. The feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal. The pressure feedback compensation designs a digital filter based on the pressure error, merges the feedforward compensation signal with the pressure feedback compensation signal, and superimposes it with the control signal to obtain the final control signal including:

[0048] Based on the combination of press input and output data acquired in advance, a press transfer function model is established. Based on different working nodes, a local linear model is established by applying Taylor series expansion, and the press transfer function parameters are determined by combining the pre-built space identification method.

[0049] Based on the press transfer function model, the zero point position and the pole position are identified, an inverse system is constructed by generating the inverse of the press transfer function, and the zero point position and the pole position are reconfigured by zero-pole configuration in combination with a low-pass filter, and discretized to obtain a discrete time model;

[0050] Based on the discrete time model, a feedforward compensation signal is calculated, and a corresponding relationship between input data and prediction results is determined, and delay compensation is performed through a feedforward prediction mechanism;

[0051] Based on the actual pressure signal obtained and the predicted expected pressure signal, the pressure error is calculated, and based on the preset adaptive filter structure, the filter coefficient is updated in real time in combination with the pressure error to obtain a pressure feedback compensation signal;

[0052] Adopting an adaptive weight allocation strategy, the feedforward compensation signal and the pressure feedback compensation signal are fused to obtain a fused compensation signal; based on the physical limit of the press, a saturation index is determined, and by detecting the saturation index, an anti-saturation algorithm is applied to correct the fused compensation signal to obtain a corrected compensation signal; the corrected compensation signal is superimposed on the control signal to generate a final control signal;

[0053] In an optional embodiment,

[0054] The feedforward compensation and the pressure feedback compensation further include:

[0055] The simulated annealing ant colony algorithm is used to automatically optimize the compensation parameters. The simulated annealing ant colony algorithm maps the compensation parameter optimization problem to the problem of ants finding the optimal path. The global optimal search is achieved through the pheromone matrix update and temperature reduction strategy to obtain the global optimal compensation parameters. The global optimal compensation parameters are applied to feedforward compensation and pressure feedback compensation, and the final control signal is generated in combination with the predictive control algorithm to achieve optimal control of the servo hydraulic system.

[0056] The simulated annealing ant colony algorithm comprises:

[0057] The compensation parameters to be optimized in feedforward compensation and pressure feedback compensation are encoded as parameter vectors, and an initial ant colony is randomly generated in the parameter feasible domain;

[0058] Based on the preset comprehensive compensation performance index, a compensation parameter objective function is determined, a simulation evaluation is performed on the position of each ant in the initial ant colony, and a compensation parameter objective function value is calculated;

[0059] Initializing a pheromone matrix based on the correlation between the compensation parameters, wherein the dimension of the pheromone matrix is ​​the same as the dimension of the parameter vector; determining attraction heuristic information based on different combinations of the compensation parameters, wherein different combinations of the compensation parameters correspond to different solutions of the ant colony;

[0060] Based on the pheromone concentration in the pheromone matrix and the attraction heuristic information, calculating the probability of each ant selecting a parameter value corresponding to the next parameter vector, and constructing a new solution according to the probability;

[0061] Perform local perturbations on the current solution of each ant to generate a perturbation solution, and calculate the compensation parameter objective function value of the perturbation solution. When the compensation parameter objective function value of the perturbation solution is greater than the compensation parameter objective function value of the current solution, accept the perturbation solution. When the compensation parameter objective function value of the perturbation solution is less than the compensation parameter objective function value of the current solution, determine the acceptance probability based on the current temperature of the simulated annealing mechanism, decide whether to accept the perturbation solution, and reduce the current temperature according to the annealing plan of the simulated annealing.

[0062] Determine and update the global optimal solution, and based on the global optimal solution, update the pheromone matrix;

[0063] Repeat the iteration until the preset number of iterations is reached, and extract the last global optimal solution as the final compensation parameter.

[0064] According to a second aspect of the embodiments of the present invention,

[0065] A servo hydraulic system optimization control system for a press test platform is provided, comprising:

[0066] The first unit is used to establish a mathematical model of the servo-hydraulic system based on the dynamic characteristics of the servo valve, hydraulic cylinder and load by deriving mathematical expressions of each component of the system and constructing a state space expression containing multiple state variables;

[0067] The second unit is used to construct a predictive control algorithm based on the mathematical model, establish a state space prediction model and determine a multi-objective optimization function, adopt a rolling optimization strategy to solve the quadratic programming problem, obtain an optimal control sequence, and use the first element of the optimal control sequence as the control input at the current moment;

[0068] The third unit is used to introduce feedforward compensation and pressure feedback compensation into the predictive control algorithm, wherein the feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, and the pressure feedback compensation designs a digital filter based on the pressure error, merges the feedforward compensation signal with the pressure feedback compensation signal, and superimposes them with the control signal to obtain the final control signal.

[0069] According to a third aspect of the embodiments of the present invention,

[0070] An electronic device is provided, comprising:

[0071] processor;

[0072] a memory for storing processor-executable instructions;

[0073] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0074] In the embodiment of the present invention, by comprehensively considering the dynamic characteristics of the servo valve, hydraulic cylinder and load, the constructed state space expression can more accurately describe the dynamic behavior of the servo hydraulic system and improve the accuracy of the system model; the mathematical model established by the state space expression can effectively support the analysis, optimization and control strategy design of the system, and improve the response performance and stability of the servo hydraulic system; by establishing a state space prediction model, the algorithm can accurately predict the future state of the system, which is helpful to formulate more accurate control strategies and improve the response speed and stability of the system; the introduction of multi-objective optimization functions allows multiple performance indicators to be considered simultaneously; through the rolling optimization strategy and the solution of the quadratic programming problem, the algorithm can find the optimal The feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, which can predict and compensate for possible errors before the system responds, thereby improving the control accuracy and tracking ability of the system; the pressure feedback compensation processes the pressure error by designing a digital filter to further reduce the actual error of the system. The feedback mechanism can correct the deviation in the pressure measurement in real time and enhance the stability and reliability of the control system; the fusion of the feedforward compensation signal and the pressure feedback compensation signal can comprehensively consider the prediction and actual feedback to form a more stable control signal, which helps to reduce the oscillation and instability of the system and improve the stability of the overall system. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is a flow chart of a servo hydraulic system optimization control method for a press test platform according to an embodiment of the present invention;

[0076] Figure 2 It is a structural schematic diagram of a servo hydraulic system optimization control system for a press test platform according to an embodiment of the present invention. DETAILED DESCRIPTION

[0077] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0078] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0079] Figure 1FIG. 1 is a flow chart of a servo hydraulic system optimization control method for a press test platform according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0080] S101. Based on the dynamic characteristics of the servo valve, hydraulic cylinder and load, a mathematical model of the servo hydraulic system is established by deriving mathematical expressions of the components of the system and constructing a state space expression containing multiple state variables;

[0081] In this embodiment, by comprehensively considering the dynamic characteristics of the servo valve, hydraulic cylinder and load, the constructed state-space expression can more accurately describe the dynamic behavior of the servo hydraulic system and improve the accuracy of the system model; the mathematical model established using the state-space expression can effectively support the analysis, optimization and control strategy design of the system and improve the response performance and stability of the servo hydraulic system.

[0082] In an optional embodiment, based on the dynamic characteristics of the servo valve, hydraulic cylinder and load, by deriving mathematical expressions of the components of the system and constructing a state space expression containing multiple state variables, the mathematical model of the servo hydraulic system is established, including:

[0083] Based on the structure and working process of the servo valve, the corresponding relationship between the valve core displacement and the flow rate of the servo valve is established, the flow equation is determined, the relationship between the valve core displacement and the control signal is constructed using a second-order vibration link model, the dynamic characteristic equation is determined, and the servo valve mathematical model is established by combining the flow equation and the dynamic characteristic equation;

[0084] Based on the relationship between the temperature change of the hydraulic oil and the bulk elastic modulus, a continuity equation is determined, the friction force is determined using the Stribeck friction model, a force balance equation is constructed, and a mathematical model of the hydraulic cylinder is established by combining the continuity equation and the force balance equation;

[0085] Based on the nonlinear characteristics of the load, nonlinear terms are generated, motion equations are constructed, and load mathematical models are established;

[0086] Determine state variables based on the servo valve mathematical model, the hydraulic cylinder mathematical model and the load mathematical model, wherein the state variables include valve core displacement, valve core speed, piston displacement, piston speed, rodless chamber pressure, rod chamber pressure, load displacement and load speed;

[0087] Establish the relationship between the time derivative of the state variable and the input quantity, generate the state equation, establish the relationship between the output quantity and the state variable and the input quantity, generate the output equation, construct the state space expression based on the state equation and the output equation, and determine the mathematical model of the servo hydraulic system accordingly.

[0088] A mathematical model is established for the servo valve. By analyzing the structure and working principle of the servo valve, observe how the valve core displacement affects the oil flow. Through experimental measurement and data fitting, the corresponding relationship between the valve core displacement and the flow is established, and the flow equation is obtained. By studying the influence of the control signal on the valve core movement and taking into account the dynamic characteristics of the valve core movement, a second-order vibration link model is used to describe the relationship between the valve core displacement and the control signal, and the dynamic characteristic equation is determined. The flow equation and the dynamic characteristic equation are combined to form a complete mathematical model of the servo valve. The second-order vibration link model specifically refers to a model used to describe a dynamic system, which uses a second-order linear differential equation to represent the dynamic behavior of the system, including the natural frequency, damping ratio and gain of the system.

[0089] Construct a mathematical model of the hydraulic cylinder. By studying the effect of changes in hydraulic oil temperature on its bulk elastic modulus, the relationship function between the two is obtained by fitting experimental data. Based on the relationship function between the two, a continuity equation is established to describe the change in oil pressure in the hydraulic cylinder cavity. Next, by analyzing the friction characteristics in the hydraulic cylinder, the Stribeck friction model is used to describe the relationship between friction and speed. Combined with the piston force analysis, a force balance equation is established. The continuity equation and the force balance equation are combined to form a complete mathematical model of the hydraulic cylinder. The Stribeck friction model specifically refers to a model for describing the change of friction with sliding speed, which is mainly used to describe the change characteristics of friction at different speeds, and is used to describe the transition behavior between low-speed friction (viscous friction) and high-speed friction (dry friction).

[0090] For the load part, first analyze the nonlinear characteristics of the load, such as stiffness nonlinearity, damping nonlinearity, etc. The nonlinear characteristics are expressed as nonlinear terms and introduced into the motion equation of the load. In this way, a mathematical model of the load including nonlinear characteristics is established.

[0091] Integrate the mathematical models of the servo valve, hydraulic cylinder, and load to determine the state variables of the system. Select the valve core displacement and valve core speed to describe the state of the servo valve; the piston displacement, piston speed, rodless chamber pressure, and rod chamber pressure to describe the state of the hydraulic cylinder; and the load displacement and load speed to describe the state of the load. All state variables together constitute the state variable set of the system.

[0092] Analyze the changing rules of each state variable, establish the relationship between the time derivative of the state variable and the system input, and obtain the state equation of the system. At the same time, determine the output of the system, and establish the relationship between the output, the state variable and the input to obtain the output equation.

[0093] The state equation and the output equation are integrated to construct the state space expression of the system. The state space expression fully describes the dynamic characteristics of the servo hydraulic system, including all the key characteristics of the servo valve, hydraulic cylinder and load, forming a complete mathematical model of the servo hydraulic system.

[0094] In this embodiment, by analyzing the structure and working principle of the servo valve, hydraulic cylinder and load, respective mathematical models are established, which helps to accurately describe the dynamic behavior of the system; a second-order vibration link model is used to describe the dynamic characteristics of the servo valve, and the Stribeck friction model is used to describe the friction force of the hydraulic cylinder, which can better capture the dynamic response of the system; the mathematical models of the servo valve, hydraulic cylinder and load are integrated, and the interaction and influence of each part are considered to provide a comprehensive understanding of the overall system behavior; by establishing state equations and output equations, the state space expression of the system is finally constructed to comprehensively describe the dynamic characteristics of the system, providing a basis for system analysis, control design and optimization; detailed mathematical models can help design more accurate control strategies, optimize the performance of the servo hydraulic system, and improve the response speed and stability of the system.

[0095] In an optional embodiment, the servo valve mathematical model includes:

[0096] The flow equation is as follows:

[0097] ;

[0098] in, Q Indicates the servo valve flow rate, K q represents the flow gain coefficient, x v Indicates the displacement of the servo valve core, Δ p Indicates the pressure difference before and after the servo valve;

[0099] The dynamic characteristic equation is as follows:

[0100] ;

[0101] in, X v ( s ) represents the Laplace transform of the servo valve spool displacement, I ( s ) represents the Laplace transform of the control signal, K i represents the current displacement gain coefficient, ω n represents the natural frequency of the servo valve, ζ represents the damping ratio, s represents a complex variable in the Laplace transform.

[0102] The flow equation describes how the flow of the servo valve is related to the valve core displacement and the pressure difference before and after the valve. In the formula, the product of the flow and the valve core displacement is proportional to the square root of the pressure difference. The flow gain coefficient is the proportional factor that determines the relationship between the flow and the valve core displacement and the pressure difference.

[0103] The dynamic characteristic equation describes the dynamic relationship between the servo valve spool displacement and the control signal. The formula shows that the relationship between the spool displacement and the control signal can be described by a second-order linear system. The dynamic behavior of this second-order system is determined by the natural frequency and the damping ratio, and the current displacement gain coefficient is the gain of the system. The equation is used to describe how the system response changes over time, especially the dynamic response characteristics of the system to the control signal.

[0104] In this embodiment, the flow equation can predict the flow of the servo valve based on the valve core displacement and the pressure difference. This helps to accurately control and regulate the fluid flow in the system and ensure that the system operates within the expected range; the dynamic characteristic equation provides the dynamic response characteristics of the servo valve to the control signal, helping to understand and optimize the dynamic behavior of the system, such as response speed, stability and transition process; the combination of the formula allows for accurate description of the performance of the servo valve, thereby supporting the design and optimization of control strategies. Through the formula, the servo control system can be better designed to improve the accuracy and efficiency of the system.

[0105] In an optional embodiment, the hydraulic cylinder mathematical model includes:

[0106] The continuity equation is as follows:

[0107] ;

[0108] in, Q 1 indicates the flow rate of the rodless cavity, Q 2 indicates the flow rate of the rod cavity. A 1 represents the effective area of ​​the rodless cavity, A 2 represents the effective area of ​​the rod cavity. x p represents the piston displacement, t Indicates time, V 1 represents the initial volume of the rodless cavity, V 2 represents the initial volume of the rod cavity, β e represents the effective bulk modulus of the liquid, p 1 indicates the rodless cavity pressure, p 2 indicates rod cavity pressure, C ip represents the internal leakage coefficient;

[0109] The force balance equation is as follows:

[0110] ;

[0111] in, m represents the equivalent mass of the piston and the load, B p represents the equivalent damping coefficient of the piston and the load, K s represents the equivalent stiffness of the hydraulic cylinder, F L Represents external load force.

[0112] The continuity equation describes the relationship between the flow change and the pressure, volume, and valve core displacement inside the hydraulic cylinder. Each equation represents the flow of the rodless chamber and the rod chamber respectively. The flow change is determined by the time derivative of the piston displacement, the pressure change of the chamber volume, and the pressure difference caused by internal leakage. It helps to describe the dynamic relationship between the flow and pressure changes of the hydraulic cylinder and supports the analysis of the fluid behavior in the hydraulic cylinder.

[0113] The force balance equation describes the force balance of the piston in the hydraulic cylinder. It relates the pressure difference between the two chambers of the hydraulic cylinder to the equation of motion of the piston. The left side of the equation represents the force generated by the pressure, while the right side represents the contribution of the piston's acceleration, velocity, displacement, and external load forces to the force balance. It helps to understand the dynamic response of the piston under different operating conditions, taking into account the mass, damping, stiffness, and external load forces of the hydraulic cylinder.

[0114] In this embodiment, the continuity equation provides the relationship between the flow and pressure, volume and piston displacement of the rodless chamber and the rod chamber, which helps to analyze and predict the fluid behavior and pressure changes inside the hydraulic cylinder, and ensure the normal operation and performance optimization of the hydraulic system; the force balance equation describes the dynamic response of the hydraulic cylinder, including the acceleration, velocity, displacement of the piston and its relationship with the hydraulic pressure difference. It is crucial to understand and optimize the motion response of the piston and ensure the stability and control accuracy of the system; the equation provides a basis for modeling the hydraulic cylinder, which can help accurately describe the behavior of the hydraulic cylinder under different operating conditions; by using the formula, the control strategy can be designed and adjusted to improve the performance of the hydraulic cylinder. Accurate flow and force balance descriptions can support more effective control and adjustment of the working state of the hydraulic cylinder, and improve the response speed and stability of the system.

[0115] In an optional embodiment, the load mathematical model includes:

[0116] The equation of motion is as follows:

[0117] ;

[0118] in, m L Indicates the load mass, xL represents the load displacement, B L represents the load damping factor, K L represents the load stiffness, F P Indicates the output force of the hydraulic cylinder.

[0119] The equation of motion describes the kinematic behavior of the load. The left side of the equation includes the load's inertial force (determined by mass and acceleration), damping force (determined by velocity), and elastic force (determined by displacement). The sum of these forces is equal to the force output by the hydraulic cylinder (right side); the inertial force is composed of the load's mass and acceleration, describing the inertia of the load due to motion; the damping force is determined by the load's velocity and damping coefficient, reflecting the resistance encountered by the load during motion; the elastic force: determined by the load's displacement and stiffness, describing the load's elastic response; the hydraulic cylinder output force: as the input force on the right side of the equation, drives the load's motion.

[0120] The equations of motion are used to describe the motion of a load under the force of a hydraulic cylinder, helping to understand and predict the dynamic behavior of the load and the effect of the hydraulic cylinder on the load.

[0121] In this embodiment, by describing the relationship between the acceleration, velocity and displacement of the load and the output force of the hydraulic cylinder, the dynamic behavior of the load can be accurately analyzed, helping to understand the response of the load under different conditions and ensure the stability and performance of the system; the formula can predict the movement of the load under the force applied by the hydraulic cylinder, providing detailed information about how the load is subjected to force and moves, which helps to optimize the system design and adjust the control strategy; the formula is used to identify and optimize the stiffness, damping and mass parameters of the load, and improve the response speed and stability of the system. It provides a theoretical basis for the design and adjustment of the load and helps improve the efficiency of the overall system.

[0122] S102. Construct a predictive control algorithm based on the mathematical model, establish a state space prediction model and determine a multi-objective optimization function, adopt a rolling optimization strategy to solve the quadratic programming problem, obtain an optimal control sequence, and use the first element of the optimal control sequence as the control input at the current moment;

[0123] In this embodiment, by establishing a state space prediction model, the algorithm can accurately predict the future state of the system, which helps to formulate a more accurate control strategy and improve the response speed and stability of the system; the introduction of multi-objective optimization functions allows multiple performance indicators to be considered simultaneously, such as control accuracy, system stability and energy efficiency. Through the rolling optimization strategy and the solution of the quadratic programming problem, the algorithm can find the optimal control sequence, thereby optimizing the overall performance of the system; the rolling optimization strategy and the use of the first element of the optimal control sequence as the control input at the current moment enable the control strategy to be adjusted in real time, and the dynamic adjustment capability ensures that the system can adapt to real-time changing working conditions and environments; by solving the quadratic programming problem and applying the optimal control sequence, the algorithm can effectively balance the control objectives and constraints, and improve the efficiency and reliability of the control strategy; the application of the predictive control algorithm helps to identify and respond to potential system instabilities, thereby enhancing the stability and robustness of the system.

[0124] In an optional embodiment, a predictive control algorithm is constructed based on the mathematical model, a state space prediction model is established and a multi-objective optimization function is determined, a rolling optimization strategy is adopted to solve the quadratic programming problem, an optimal control sequence is obtained, and the first element of the optimal control sequence is used as the control input at the current moment, including:

[0125] Based on the mathematical model, a continuous-time state-space expression is determined, and a discrete-time state-space model is obtained after discretization.

[0126] Based on the discrete time state space model, according to the preset prediction time domain, the state of multiple future time steps is calculated, the output prediction equation is derived, the prediction result is obtained, and the prediction result is decomposed into a free response and a forced response;

[0127] According to the preset control target, based on the free response and the forced response, a multi-objective optimization function including a state error index, a control energy index and a control increment index is generated; based on the predetermined physical characteristics and the preset control requirements, based on the discrete time state space model and the output prediction equation, constraint conditions including control input constraints, control increment constraints, state variable constraints and output variable constraints are determined;

[0128] Expand the multi-objective optimization function into a quadratic form, generate state error terms, control input terms and control increment terms, and construct a Hessian matrix and a gradient vector in combination with the state control cross terms that determine the coupling relationship, and determine the objective function of the quadratic programming problem; based on the constraint conditions, determine linear inequality constraints and linear equality constraints, and generate a constraint matrix and a constraint vector in combination with the state prediction matrix and the control matrix generated based on the discrete time state space model and the prediction time domain, and standardize the quadratic programming problem;

[0129] Based on the preprocessing of the Hessian matrix through the Cholesky decomposition method, the optimal solution of the previous time step is rolled to the current time step as the starting solution, a new state is obtained, the discrete-time state space model is updated, and the standardized quadratic programming problem is iteratively solved to obtain the optimal control sequence.

[0130] Firstly, based on the established servo hydraulic system state space mathematical model, the continuous time state space expression is determined. Since the actual control system is implemented in a discrete time form in a computer, it is necessary to discretize the continuous time state space expression. The continuous time state space expression is discretized using a zero-order holder to obtain a discrete time state space model, which describes the dynamic evolution relationship of the system state variables at discrete time points.

[0131] Based on the obtained discrete time state space model, the state of multiple future time steps is calculated according to the preset prediction time domain. Through iterative calculation, the predicted values ​​of the state variables at each time step in the future prediction time domain from the current moment can be obtained. Based on the predicted values ​​of the state variables, the output prediction equation is derived to obtain the predicted values ​​of the output variables at each time step in the future prediction time domain. In order to facilitate subsequent optimization calculations, the obtained predicted values ​​of the state variables and the predicted values ​​of the output variables are decomposed into free response and forced response. Among them, the free response represents the response of the system under zero input conditions, and the forced response represents the response of the system under given input conditions.

[0132] According to the preset control objectives, based on the obtained free response and forced response, a multi-objective optimization function including state error index, control energy index and control increment index is generated. The state error index is used to measure the deviation between the system state variable and the expected value, the control energy index is used to measure the energy consumption of the control input, and the control increment index is used to measure the change of the control input between two adjacent time steps. By reasonably selecting the weight coefficient of each indicator, the dynamic performance and control cost of the system can be balanced. At the same time, based on the physical characteristics of the servo hydraulic system and the preset control requirements, for example, such as the displacement limit, speed limit, pressure limit, etc. of the actuator, based on the discrete time state space model and the output prediction equation, the constraints including control input constraints, control increment constraints, state variable constraints and output variable constraints are determined to ensure that the system operates within the physically allowed range.

[0133] In order to transform the multi-objective optimization problem into a standard form that is easy to solve, the generated multi-objective optimization function is expanded into a quadratic form. By combining similar terms and replacing variables, a quadratic objective function consisting of state error terms, control input terms, control increment terms, and state control cross terms can be obtained. By organizing the quadratic objective function, the Hessian matrix and gradient vector can be constructed to determine the objective function of the quadratic programming problem. At the same time, based on the determined constraints, equivalent linear inequality constraints and linear equality constraints can be obtained through mathematical transformation. Combined with the state prediction matrix and control matrix generated based on the discrete time state space model and the prediction domain, the constraint matrix and constraint vector corresponding to the linear constraints can be generated, thereby standardizing the multi-objective optimization problem into a quadratic programming problem.

[0134] An effective numerical optimization algorithm is used to solve the standardized quadratic programming problem and obtain the optimal control sequence. In order to improve the computational efficiency and numerical stability, the Hessian matrix is ​​preprocessed by the Cholesky decomposition method before solving, and it is decomposed into a form that is easy to invert. At the same time, in order to speed up the convergence of the optimization problem, the optimal solution of the previous time step is used as the starting solution of the optimization problem of the current time step. At each time step, the discrete time state space model is updated according to the actual state of the system, and the quadratic programming problem is iteratively solved until the preset convergence condition is reached, and the optimal control quantity of the current time step can be obtained. The optimal control quantity is input into the servo hydraulic system as a control instruction, and rolling optimization is carried out to realize the model predictive control of the servo hydraulic system.

[0135] In this embodiment, by discretizing the continuous-time state space model and applying it to the prediction time domain, the dynamic behavior of the system at discrete time points can be accurately described, providing a basis for the state and output prediction of multiple time steps in the future, ensuring the accuracy of the prediction; based on the generated multi-objective optimization function (including state error, control energy and control increment index), the dynamic performance and control cost of the system can be balanced, ensuring that the system controls energy consumption and control input changes while meeting the control objectives; in the optimization process, physical constraints (such as displacement, speed and pressure limits) are taken into account to ensure that the system does not exceed physical limitations in actual operation, which helps to achieve safe and reliable control; the multi-objective optimization problem is converted into a standard quadratic programming problem, and mathematical processing (such as Cholesky decomposition of the Hessian matrix) is used to improve computational efficiency and numerical stability, making the solution process more efficient and stable; through rolling optimization and real-time updating of the state space model, the optimization problem is iteratively solved according to the current system state to ensure that the control strategy can be adjusted and adapted to the dynamic changes of the system in real time.

[0136] S103. In the predictive control algorithm, feedforward compensation and pressure feedback compensation are introduced, wherein the feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, and the pressure feedback compensation designs a digital filter based on the pressure error, merges the feedforward compensation signal with the pressure feedback compensation signal, and superimposes them with the control signal to obtain the final control signal.

[0137] In this embodiment, feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, which can predict and compensate for possible errors before the system responds, thereby improving the control accuracy and the tracking ability of the system; pressure feedback compensation processes the pressure error by designing a digital filter to further reduce the actual error of the system, and the feedback mechanism can correct the deviation in the pressure measurement in real time, enhancing the stability and reliability of the control system; the feedforward compensation signal is integrated with the pressure feedback compensation signal, which can comprehensively consider the prediction and actual feedback to form a more stable control signal, which helps to reduce the oscillation and instability of the system and improve the stability of the overall system; feedforward compensation can quickly respond to system changes and reduce system delays. Combined with pressure feedback compensation, it can optimize the response speed and dynamic performance of the system and make the system more sensitive to disturbances and changes; the feedforward and feedback signals are superimposed to form the final control signal, which can comprehensively consider the prediction and actual feedback information of the system, thereby achieving more comprehensive control compensation and improving the overall performance of the system.

[0138] In an optional embodiment, feedforward compensation and pressure feedback compensation are introduced into the predictive control algorithm, wherein the feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, and the pressure feedback compensation designs a digital filter based on the pressure error, fuses the feedforward compensation signal with the pressure feedback compensation signal, and superimposes the signal with the control signal to obtain a final control signal including:

[0139] Based on the combination of press input and output data acquired in advance, a press transfer function model is established. Based on different working nodes, a local linear model is established by applying Taylor series expansion, and the press transfer function parameters are determined by combining the pre-built space identification method.

[0140] Based on the press transfer function model, the zero point position and the pole position are identified, an inverse system is constructed by generating the inverse of the press transfer function, and the zero point position and the pole position are reconfigured by zero-pole configuration in combination with a low-pass filter, and discretized to obtain a discrete time model;

[0141] Based on the discrete time model, a feedforward compensation signal is calculated, and a corresponding relationship between input data and prediction results is determined, and delay compensation is performed through a feedforward prediction mechanism;

[0142] Based on the actual pressure signal obtained and the predicted expected pressure signal, the pressure error is calculated, and based on the preset adaptive filter structure, the filter coefficient is updated in real time in combination with the pressure error to obtain a pressure feedback compensation signal;

[0143] An adaptive weight allocation strategy is adopted to fuse the feedforward compensation signal and the pressure feedback compensation signal to obtain a fused compensation signal; based on the physical limit of the press, a saturation index is determined, and by detecting the saturation index and applying an anti-saturation algorithm, the fused compensation signal is corrected to obtain a corrected compensation signal; the corrected compensation signal is superimposed on the control signal to generate a final control signal.

[0144] Based on the pre-acquired input and output data combination of the press, the transfer function model of the press is established by using the system identification technology. Considering the differences in the dynamic characteristics of the press at different working nodes, the Taylor series expansion method is applied to different working nodes to establish a local linear model near the working node. In order to accurately determine the parameters of the transfer function model of the press, the optimal transfer function parameters are obtained by optimization calculation combined with the pre-built spatial identification method.

[0145] Based on the established transfer function model of the press, the zero and pole positions of the press system are identified by analyzing the zero-pole distribution of the transfer function. In order to achieve precise control of the press, an inverse system is constructed by generating the inverse of the press transfer function to offset the dynamic characteristics of the press system. At the same time, in order to ensure the stability and robustness of the inverse system, the zero and pole positions of the inverse system are reconfigured through the zero-pole configuration technology, and a low-pass filter is introduced to suppress high-frequency interference. Considering the implementation of the control algorithm in the digital controller, the designed continuous-time inverse system model is discretized to obtain a discrete-time model.

[0146] Based on the obtained discrete time model, the feedforward compensation signal is calculated. By analyzing the dynamic characteristics of the press system and the expected pressure output, the corresponding relationship between the input data and the prediction results is determined. Using this corresponding relationship, the input data at the future moment is calculated in advance through the feedforward prediction mechanism and used as the feedforward compensation signal to compensate for the time delay of the system.

[0147] At the same time, in order to further improve the control accuracy and adaptability, a feedback control mechanism is introduced. The pressure error is calculated by obtaining the actual pressure signal of the press and the predicted expected pressure signal. Based on the preset adaptive filter structure, preferably, the adaptive LMS filter, combined with the calculated pressure error, updates the filter coefficient in real time to adapt to the dynamic changes of the press system. The pressure error is processed by the adaptive filter to obtain a pressure feedback compensation signal.

[0148] In order to comprehensively utilize the advantages of feedforward compensation and feedback compensation, an adaptive weight allocation strategy is adopted to fuse the calculated feedforward compensation signal with the pressure feedback compensation signal. By reasonably setting the weights of feedforward compensation and feedback compensation, the proportional relationship between them is dynamically adjusted to obtain a fused compensation signal. Considering the physical limits of the press actuator, a saturation index is introduced. By real-time detection of the saturation index, it is determined whether the fused compensation signal exceeds the physical limit. When saturation is detected, the anti-saturation algorithm is applied to correct the fused compensation signal so that it meets the physical constraints and obtains the corrected compensation signal.

[0149] The corrected compensation signal is superimposed on the control signal of the press to generate the final control signal. The generated control signal is input into the press system to drive the press actuator to move and realize precise control of the press.

[0150] In this embodiment, by establishing a local linear model of the press, considering the dynamic characteristics of the press at different working nodes, providing an accurate transfer function model, ensuring the accuracy of system modeling; using the inverse system and zero-pole configuration technology of the press transfer function, a stable and robust inverse system is constructed to successfully offset the dynamic characteristics of the press, improve the control accuracy and system response speed; by calculating the feedforward compensation signal and using the feedforward prediction mechanism to compensate for the delay effect of the system, the early adjustment of future input data is achieved, which helps to reduce the negative impact of the system delay on the control accuracy and improve the instant response capability of the system; introducing an adaptive filter to process the pressure error to achieve dynamic adjustment and optimized feedback control. The adaptive filter updates the filter coefficient according to the real-time pressure error to improve the adaptability and accuracy of the control system; through the adaptive weight allocation strategy, the feedforward compensation and feedback compensation signals are fused, making full use of the advantages of the two compensation methods and enhancing the control effect of the system; introducing saturation indicators and anti-saturation algorithms to ensure that the fused compensation signal does not exceed the physical limit, so that the control signal can still effectively achieve accurate control of the press while meeting the physical constraints.

[0151] In an optional embodiment, the feedforward compensation and the pressure feedback compensation further include:

[0152] The simulated annealing ant colony algorithm is used to automatically optimize the compensation parameters. The simulated annealing ant colony algorithm maps the compensation parameter optimization problem to the problem of ants finding the optimal path. The global optimal search is achieved through the pheromone matrix update and temperature reduction strategy to obtain the global optimal compensation parameters. The global optimal compensation parameters are applied to feedforward compensation and pressure feedback compensation, and the final control signal is generated in combination with the predictive control algorithm to achieve optimal control of the servo hydraulic system.

[0153] The simulated annealing ant colony algorithm comprises:

[0154] The compensation parameters to be optimized in feedforward compensation and pressure feedback compensation are encoded as parameter vectors, and an initial ant colony is randomly generated in the parameter feasible domain;

[0155] Based on the preset comprehensive compensation performance index, a compensation parameter objective function is determined, a simulation evaluation is performed on the position of each ant in the initial ant colony, and a compensation parameter objective function value is calculated;

[0156] Initializing a pheromone matrix based on the correlation between the compensation parameters, wherein the dimension of the pheromone matrix is ​​the same as the dimension of the parameter vector; determining attraction heuristic information based on different combinations of the compensation parameters, wherein different combinations of the compensation parameters correspond to different solutions of the ant colony;

[0157] Based on the pheromone concentration in the pheromone matrix and the attraction heuristic information, calculating the probability of each ant selecting a parameter value corresponding to the next parameter vector, and constructing a new solution according to the probability;

[0158] Perform local perturbations on the current solution of each ant to generate a perturbation solution, and calculate the compensation parameter objective function value of the perturbation solution. When the compensation parameter objective function value of the perturbation solution is greater than the compensation parameter objective function value of the current solution, accept the perturbation solution. When the compensation parameter objective function value of the perturbation solution is less than the compensation parameter objective function value of the current solution, determine the acceptance probability based on the current temperature of the simulated annealing mechanism, decide whether to accept the perturbation solution, and reduce the current temperature according to the annealing plan of the simulated annealing.

[0159] Determine and update the global optimal solution, and based on the global optimal solution, update the pheromone matrix;

[0160] Repeat the iteration until the preset number of iterations is reached, and extract the last global optimal solution as the final compensation parameter.

[0161] The compensation parameters to be optimized in feedforward compensation and pressure feedback compensation are encoded as parameter vectors to form the solution space of the ant colony algorithm. In the feasible domain of parameters, the initial ant colony is randomly generated, and each ant represents a set of candidate solutions for the compensation parameters to be optimized.

[0162] Based on the preset comprehensive compensation performance indicators, such as system response time, overshoot, steady-state error, etc., the compensation parameter objective function is constructed as the basis for evaluating the quality of the compensation parameters. For each ant in the initial ant colony, the compensation parameters corresponding to its position are substituted into the servo hydraulic system model for simulation, and the objective function value of the group of compensation parameters is calculated through the simulation results as the fitness value of the ant.

[0163] Considering the correlation between different compensation parameters, a pheromone matrix with the same dimension as the parameter vector is initialized to record the experience information accumulated during the ant colony search process. At the same time, based on different combinations of compensation parameters, that is, different solutions of the ant colony, the attraction heuristic information is determined to represent the relative advantages and disadvantages between different solutions.

[0164] During the ant colony search process, the probability of each ant selecting the parameter value corresponding to the next parameter vector is calculated based on the pheromone concentration and attraction heuristic information in the pheromone matrix. According to the calculated probability, the ant constructs a new solution through a selection mechanism such as roulette, that is, generates a new compensation parameter combination.

[0165] In order to increase the randomness of the ant colony search and the ability to jump out of the local optimum, the current solution of each ant is locally perturbed to generate a perturbed solution. The compensation parameter objective function value of the perturbed solution is calculated and compared with the objective function value of the current solution. If the objective function value of the perturbed solution is greater than the current solution, the perturbed solution is accepted as the new current solution; if the objective function value of the perturbed solution is less than the current solution, an acceptance probability is calculated based on the current temperature of the simulated annealing mechanism, and the decision on whether to accept the inferior perturbation solution is made based on the probability. At the same time, according to the preset annealing plan, the current temperature is gradually lowered, so that the algorithm accepts inferior solutions with a higher probability in the early stage, expands the search range, and accepts inferior solutions with a lower probability in the later stage, accelerating the convergence to the global optimum.

[0166] In each iteration, the current solutions of all ants are compared, and the global optimal solution, that is, the compensation parameter combination with the largest objective function value, is updated and recorded. Based on the global optimal solution, the pheromone matrix is ​​updated to enhance the pheromone concentration around the high-quality solution and guide the ant colony to move in a promising search direction.

[0167] Repeat the above ant colony search and simulated annealing process until the preset number of iterations is reached. The global optimal solution obtained in the last iteration is extracted as the optimal combination of compensation parameters.

[0168] The optimized compensation parameters are applied to feedforward compensation and pressure feedback compensation, and combined with the predictive control algorithm to generate the final control signal of the servo hydraulic system to achieve optimal control of the system.

[0169] In this embodiment, by encoding the compensation parameter optimization problem into the solution space of the ant colony algorithm and utilizing the global search capability of the ant colony algorithm, the problem that the traditional optimization method is prone to fall into the local optimum is effectively avoided, and the probability of finding the global optimal compensation parameter combination is improved; by combining the pheromone matrix and heuristic information, the ant colony can adaptively search for the optimal solution in the solution space, and as the iterative process proceeds, the high-quality solution is gradually concentrated, thereby improving the optimization efficiency; by using the simulated annealing mechanism for local perturbations, the algorithm has a certain degree of randomness and exploration ability, and can effectively jump out of the local optimal solution, further improving the success rate of global optimization; by gradually reducing the temperature of the annealing plan, the optimization algorithm can search extensively and expand the solution space in the early stage, and concentrate on searching and accelerating convergence in the later stage, thereby improving the dynamic adaptability and convergence speed of the algorithm; by combining the optimized compensation parameters with the predictive control algorithm, a more accurate control signal is generated, which significantly improves the control accuracy, response speed and stability of the servo hydraulic system, and realizes the optimal control of the system.

[0170] Figure 2 FIG. 1 is a schematic diagram of a structure of a servo hydraulic system optimization control system for a press test platform according to an embodiment of the present invention. Figure 2 As shown, the system comprises:

[0171] The first unit is used to establish a mathematical model of the servo-hydraulic system based on the dynamic characteristics of the servo valve, hydraulic cylinder and load by deriving mathematical expressions of each component of the system and constructing a state space expression containing multiple state variables;

[0172] The second unit is used to construct a predictive control algorithm based on the mathematical model, establish a state space prediction model and determine a multi-objective optimization function, adopt a rolling optimization strategy to solve the quadratic programming problem, obtain an optimal control sequence, and use the first element of the optimal control sequence as the control input at the current moment;

[0173] The third unit is used to introduce feedforward compensation and pressure feedback compensation into the predictive control algorithm, wherein the feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, and the pressure feedback compensation designs a digital filter based on the pressure error, merges the feedforward compensation signal with the pressure feedback compensation signal, and superimposes them with the control signal to obtain the final control signal.

[0174] According to a third aspect of the embodiments of the present invention,

[0175] An electronic device is provided, comprising:

[0176] processor;

[0177] a memory for storing processor-executable instructions;

[0178] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0179] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0180] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A servo hydraulic system optimization control method for a press test platform, characterized in that: include: Based on the dynamic characteristics of the servo valve, hydraulic cylinder and load, the mathematical model of the servo hydraulic system is established by deriving the mathematical expressions of each component of the system and constructing a state space expression containing multiple state variables; Based on the mathematical model, a predictive control algorithm is constructed, by establishing a state space prediction model and determining a multi-objective optimization function, a rolling optimization strategy is adopted to solve the quadratic programming problem, an optimal control sequence is obtained, and the first element of the optimal control sequence is used as the control input at the current moment; Feedforward compensation and pressure feedback compensation are introduced into the predictive control algorithm, wherein the feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, and the pressure feedback compensation designs a digital filter based on the pressure error, fuses the feedforward compensation signal with the pressure feedback compensation signal, and superimposes the signal with the control signal to obtain the final control signal; The feedforward compensation and the pressure feedback compensation further include: The simulated annealing ant colony algorithm is used to automatically optimize the compensation parameters. The simulated annealing ant colony algorithm maps the compensation parameter optimization problem to the problem of ants finding the optimal path. The global optimal search is achieved through the pheromone matrix update and temperature reduction strategy to obtain the global optimal compensation parameters. The global optimal compensation parameters are applied to feedforward compensation and pressure feedback compensation, and the final control signal is generated in combination with the predictive control algorithm to achieve optimal control of the servo hydraulic system. The simulated annealing ant colony algorithm comprises: The compensation parameters to be optimized in feedforward compensation and pressure feedback compensation are encoded as parameter vectors, and an initial ant colony is randomly generated in the parameter feasible domain; Based on the preset comprehensive compensation performance index, a compensation parameter objective function is determined, a simulation evaluation is performed on the position of each ant in the initial ant colony, and a compensation parameter objective function value is calculated; Initializing a pheromone matrix based on the correlation between the compensation parameters, wherein the dimension of the pheromone matrix is ​​the same as the dimension of the parameter vector; determining attraction heuristic information based on different combinations of the compensation parameters, wherein different combinations of the compensation parameters correspond to different solutions of the ant colony; Based on the pheromone concentration in the pheromone matrix and the attraction heuristic information, calculating the probability of each ant selecting a parameter value corresponding to the next parameter vector, and constructing a new solution according to the probability; Perform local perturbations on the current solution of each ant to generate a perturbation solution, and calculate the compensation parameter objective function value of the perturbation solution. When the compensation parameter objective function value of the perturbation solution is greater than the compensation parameter objective function value of the current solution, accept the perturbation solution. When the compensation parameter objective function value of the perturbation solution is less than the compensation parameter objective function value of the current solution, determine the acceptance probability based on the current temperature of the simulated annealing mechanism, decide whether to accept the perturbation solution, and reduce the current temperature according to the annealing plan of the simulated annealing. Determine and update the global optimal solution, and based on the global optimal solution, update the pheromone matrix; Repeat the iteration until the preset number of iterations is reached, and extract the last global optimal solution as the final compensation parameter.

2. The method according to claim 1, characterized in that: Based on the dynamic characteristics of the servo valve, hydraulic cylinder and load, the mathematical expressions of the system components are derived, and the state space expression containing multiple state variables is constructed. The mathematical model of the servo hydraulic system is established, including: Based on the structure and working process of the servo valve, the corresponding relationship between the valve core displacement and the flow rate of the servo valve is established, the flow equation is determined, the relationship between the valve core displacement and the control signal is constructed using a second-order vibration link model, the dynamic characteristic equation is determined, and the servo valve mathematical model is established by combining the flow equation and the dynamic characteristic equation; Based on the relationship between the temperature change of the hydraulic oil and the bulk elastic modulus, a continuity equation is determined, the friction force is determined using the Stribeck friction model, a force balance equation is constructed, and a mathematical model of the hydraulic cylinder is established by combining the continuity equation and the force balance equation; Based on the nonlinear characteristics of the load, nonlinear terms are generated, motion equations are constructed, and load mathematical models are established; Determine state variables based on the servo valve mathematical model, the hydraulic cylinder mathematical model and the load mathematical model, wherein the state variables include valve core displacement, valve core speed, piston displacement, piston speed, rodless chamber pressure, rod chamber pressure, load displacement and load speed; Establish the relationship between the time derivative of the state variable and the input quantity, generate the state equation, establish the relationship between the output quantity and the state variable and the input quantity, generate the output equation, construct the state space expression based on the state equation and the output equation, and determine the mathematical model of the servo hydraulic system accordingly.

3. The method according to claim 2, characterized in that The servo valve mathematical model includes: The flow equation is as follows: ； in, Q Indicates the servo valve flow rate, K q represents the flow gain coefficient, x v Indicates the displacement of the servo valve core, Δ p Indicates the pressure difference before and after the servo valve; The dynamic characteristic equation is as follows: ； in, X v ( s ) represents the Laplace transform of the servo valve spool displacement, I ( s ) represents the Laplace transform of the control signal, K i represents the current displacement gain coefficient, ω n represents the natural frequency of the servo valve, ζ represents the damping ratio, s represents a complex variable in the Laplace transform.

4. The method according to claim 2, characterized in that: The hydraulic cylinder mathematical model includes: The continuity equation is as follows: ； in, Q 1 indicates the flow rate of the rodless cavity, Q 2 indicates the flow rate of the rod cavity. A 1 represents the effective area of ​​the rodless cavity, A 2 represents the effective area of ​​the rod cavity. x p represents the piston displacement, t Indicates time, V 1 represents the initial volume of the rodless cavity, V 2 represents the initial volume of the rod cavity, β e represents the effective bulk modulus of the liquid, p 1 indicates the rodless cavity pressure, p 2 indicates rod cavity pressure, C ip represents the internal leakage coefficient; The force balance equation is as follows: ； in, m represents the equivalent mass of the piston and the load, B p represents the equivalent damping coefficient of the piston and the load, K s represents the equivalent stiffness of the hydraulic cylinder, F L Represents external load force.

5. The method according to claim 2, characterized in that: The load mathematical model includes: The equation of motion is as follows: ； in, m L Indicates the load mass, x L represents the load displacement, B L represents the load damping factor, K L represents the load stiffness, F P Indicates the output force of the hydraulic cylinder.

6. The method according to claim 1, characterized in that Based on the mathematical model, a predictive control algorithm is constructed. By establishing a state space prediction model and determining a multi-objective optimization function, a rolling optimization strategy is used to solve the quadratic programming problem, and the optimal control sequence is obtained. The first element of the optimal control sequence is used as the control input at the current moment, including: Based on the mathematical model, a continuous-time state-space expression is determined, and a discrete-time state-space model is obtained after discretization. Based on the discrete time state space model, according to the preset prediction time domain, the state of multiple time steps in the future is calculated, the output prediction equation is derived, the prediction result is obtained, and the prediction result is decomposed into a free response and a forced response; According to the preset control target, based on the free response and the forced response, a multi-objective optimization function including a state error index, a control energy index and a control increment index is generated; based on the predetermined physical characteristics and the preset control requirements, based on the discrete time state space model and the output prediction equation, constraint conditions including control input constraints, control increment constraints, state variable constraints and output variable constraints are determined; Expand the multi-objective optimization function into a quadratic form, generate state error terms, control input terms and control increment terms, and construct a Hessian matrix and a gradient vector in combination with the state control cross terms that determine the coupling relationship, and determine the objective function of the quadratic programming problem; based on the constraint conditions, determine linear inequality constraints and linear equality constraints, and generate a constraint matrix and a constraint vector in combination with the state prediction matrix and the control matrix generated based on the discrete time state space model and the prediction time domain, and standardize the quadratic programming problem; Based on the preprocessing of the Hessian matrix through the Cholesky decomposition method, the optimal solution of the previous time step is rolled to the current time step as the starting solution, a new state is obtained, the discrete-time state space model is updated, and the standardized quadratic programming problem is iteratively solved to obtain the optimal control sequence.

7. The method according to claim 1, characterized in that Feedforward compensation and pressure feedback compensation are introduced into the predictive control algorithm. The feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal. The pressure feedback compensation designs a digital filter based on the pressure error, merges the feedforward compensation signal with the pressure feedback compensation signal, and superimposes it with the control signal to obtain the final control signal including: Based on the combination of press input and output data acquired in advance, a press transfer function model is established. Based on different working nodes, a local linear model is established by applying Taylor series expansion, and the press transfer function parameters are determined by combining the pre-built space identification method. Based on the press transfer function model, the zero point position and the pole position are identified, an inverse system is constructed by generating the inverse of the press transfer function, and the zero point position and the pole position are reconfigured by zero-pole configuration in combination with a low-pass filter, and discretized to obtain a discrete time model; Based on the discrete time model, a feedforward compensation signal is calculated, and a corresponding relationship between input data and prediction results is determined, and delay compensation is performed through a feedforward prediction mechanism; Based on the actual pressure signal obtained and the predicted expected pressure signal, the pressure error is calculated, and based on the preset adaptive filter structure, the filter coefficient is updated in real time in combination with the pressure error to obtain a pressure feedback compensation signal; An adaptive weight allocation strategy is adopted to fuse the feedforward compensation signal and the pressure feedback compensation signal to obtain a fused compensation signal; based on the physical limit of the press, a saturation index is determined, and by detecting the saturation index and applying an anti-saturation algorithm, the fused compensation signal is corrected to obtain a corrected compensation signal; the corrected compensation signal is superimposed on the control signal to generate a final control signal.

8. A servo hydraulic system optimization control system for a press test platform, used to implement the method described in any one of claims 1 to 7, characterized in that: include: The first unit is used to establish a mathematical model of the servo-hydraulic system based on the dynamic characteristics of the servo valve, hydraulic cylinder and load by deriving mathematical expressions of each component of the system and constructing a state space expression containing multiple state variables; The second unit is used to construct a predictive control algorithm based on the mathematical model, establish a state space prediction model and determine a multi-objective optimization function, adopt a rolling optimization strategy to solve the quadratic programming problem, obtain an optimal control sequence, and use the first element of the optimal control sequence as the control input at the current moment; The third unit is used to introduce feedforward compensation and pressure feedback compensation into the predictive control algorithm, wherein the feedforward compensation constructs a compensation model based on the inverse system of the press transfer function to pre-correct the control signal, and the pressure feedback compensation designs a digital filter based on the pressure error, merges the feedforward compensation signal with the pressure feedback compensation signal, and superimposes them with the control signal to obtain the final control signal.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 7.

Images