An aviation precision forging forging and pressing processing adaptive control system and method

Abstract

The application relates to the field of metal plastic forming control, and discloses an aviation precision forging forging and pressing processing adaptive control system and method, which comprises the following: an online dynamic model identification module which identifies a process fingerprint representing current blank characteristics in an initial processing stage; a target curve adaptive regulator which adjusts a reference energy dissipation rate curve according to the fingerprint to generate a personalized reference curve; a model predictive controller which uses continuously updated model parameters to perform rolling optimization with the reference curve as a tracking target and outputs optimal speed instructions; and an event triggering and constraint dynamic modulation module which actively tightens operation constraints applied to the controller by monitoring model prediction errors in parallel and before predicting a sudden increase in forging pressure event. The application realizes control over the forging and pressing process by combining target adaptation and constraint dynamic modulation, can effectively suppress forging pressure overshoot, and significantly improves the forming precision of forgings and the batch-to-batch quality consistency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of metal plastic forming control, in particular to a forging and pressing processing adaptive control system and method for aviation precision forgings. BACKGROUND

[0002] Forging and pressing is the core process of material precision forming, and the product quality directly affects the performance and service life of high-end equipment. Especially in the field of aerospace, for key load-bearing forgings such as turbine discs and casings made of titanium alloys and high-temperature alloys, the microstructure and performance and dimensional accuracy are extremely demanding. Precise control of the forging and pressing process is the key to ensuring the stability of the quality of forgings.

[0003] Currently, the control mode of forging and pressing equipment mostly relies on pre-set process programs. The controller executes the processing according to a fixed punch speed or forging pressure curve. This method can complete the basic forming task when facing ideal and constant processing conditions. However, in actual industrial production, there is a deep contradiction between the rigidity of this control strategy and the high uncertainty of the process. On the one hand, the billets entering the furnace inevitably fluctuate in initial physical properties, such as small differences in chemical composition, unevenness of heating temperature, or heat loss during transfer, which will cause the deformation resistance to deviate from the nominal value. The solidified control program cannot perceive and adapt to the differences between billets, resulting in significant fluctuations in the forming quality of the final forgings between different batches.

[0004] On the other hand, even for a single billet, forging and pressing itself is a dramatic nonlinear process with constantly changing physical properties. Within milliseconds of the punch pressing, multiple complex physical and metallurgical phenomena such as work hardening, dynamic recovery and recrystallization, and deformation heat effects occur simultaneously inside the material, which will cause the deformation resistance to present complex dynamic changes at different stages of the processing stroke. Traditional control methods, even if PID closed-loop control is used, are difficult to achieve continuous optimal tracking and adjustment of this dynamic process throughout the stroke.

[0005] The more severe challenge occurs in the final stage of the forging process. When the metal is about to fully fill the die cavity, the flow space of the material is sharply reduced, and the equivalent deformation resistance will show an exponential sharp increase, thus generating a huge forging pressure in a very short time. The response of the existing control system often has a lag, and they can only passively adjust after monitoring that the pressure exceeds the threshold, but at this time the pressure growth rate is very fast enough to make the actual forging pressure seriously exceed the carrying capacity of the equipment or the die. This pressure overshoot not only causes damage to the expensive die and forging press body, shortening its service life, but also seriously affects the final size accuracy and internal organization state of the forged piece, which is a technical problem that has been long been solved in the field of precision forging. Therefore, developing an advanced control method that can actively adapt to the changes in working conditions and foreseeably avoid process risks is of great significance to improve the core competitiveness of China's high-end forging manufacturing industry. SUMMARY

[0006] The technical problem solved by the present application is that the existing aviation precision forging forging process control method usually uses fixed process parameter program, which is difficult to adapt to the changes brought by the uncertainty factors such as the initial state of the blank, the lubrication condition and the die temperature in the processing process, thereby leading to low consistency of the finished forging quality, and it is difficult to optimize the control of the forging process.

[0007] To solve the above technical problems, the present application provides an aviation precision forging forging process self-adaptive control system.

[0008] The first aspect of the present application provides an aviation precision forging forging process self-adaptive control system.

[0009] The self-adaptive control system comprises an online dynamic model identification module, a model predictive controller and an event trigger and constraint dynamic modulation module.

[0010] The online dynamic model identification module is configured to collect the forging pressure, punch speed and punch displacement data in real time through the connected sensors during the forging process. The online dynamic model identification module updates the model parameters of a lumped parameter physical model online and periodically based on the collected data using a preset identification algorithm. The lumped parameter physical model is used to characterize the dynamic relationship between the forging pressure and the punch speed and the punch displacement.

[0011] In one specific embodiment, the lumped parameter physical model can have the following form:

[0012] ;

[0013] wherein, is the forging pressure, is the punch displacement, is the punch speed, The strain rate sensitivity index of the material, the model parameters include the equivalent stiffness coefficient. and equivalent viscous damping coefficient .

[0014] The model predictive controller is communicatively connected to the online dynamic model identification module. The model predictive controller is configured to receive the model parameters, which are updated in real-time and output by the online dynamic model identification module. The model predictive controller also stores a preset reference energy dissipation rate curve. Based on the received model parameters and the reference energy dissipation rate curve, the model predictive controller calculates control commands for controlling the punch speed within a finite prediction time domain by solving a constrained optimization problem, and outputs the control commands to the actuator of the forging press.

[0015] In one specific embodiment, the optimization problem solved by the model predictive controller includes the following objective function:

[0016] ;

[0017] in, This is the index of the current discrete time step; To be in discrete time steps Calculate the objective function value; To predict the length of the time domain, To control the length of the time domain; To be in discrete time steps For the future Predicted energy dissipation rate for each discrete time step; The reference energy dissipation rate curve is used in the future... The target value for each discrete time step; To be in discrete time steps Calculated, for the future i-th The punch speed increment to be optimized at discrete time steps; This is the weight matrix used to penalize tracking errors in energy dissipation rate; This is a weight matrix used to suppress the incremental change in the punch speed.

[0018] The event-triggered and constraint dynamic modulation module is connected with the online dynamic model identification module and the model predictive controller. The event-triggered and constraint dynamic modulation module is configured to calculate, in real time, a prediction error between a predicted value of a model used by the online dynamic model identification module and an actual value collected by a sensor. The event-triggered and constraint dynamic modulation module continuously monitors a variation trend of the prediction error. When it is monitored that the prediction error presents a preset abnormal trend (for example, continuously increasing in one direction), the event-triggered and constraint dynamic modulation module outputs a modulation signal to the model predictive controller, and the modulation signal is used to dynamically adjust a constraint condition in an optimization problem of the model predictive controller, for example, tightening a maximum forging pressure constraint. This design enables the system to predict the occurrence of a key physical event such as the end of die filling and to adjust the control strategy in advance to ensure process stability.

[0019] Further, the adaptive control system can further include a target curve adaptive adjuster. The target curve adaptive adjuster is connected with the online dynamic model identification module and the model predictive controller. The target curve adaptive adjuster is configured to, in an initial stage of the forging process, acquire initial model parameters identified by the online dynamic model identification module and take the initial model parameters as a process fingerprint of a current workpiece. Subsequently, the target curve adaptive adjuster performs mathematical operation adjustment on a preset reference energy dissipation rate curve according to the process fingerprint to generate a new curve, and takes the newly generated curve as the reference energy dissipation rate curve for the model predictive controller. In this way, the control target can be adaptively adjusted according to the initial physical characteristics of each billet.

[0020] The second aspect of the present application provides an adaptive control method for aviation precision forging.

[0021] The method includes the following steps:

[0022] a) In an initial stage of the forging process, initial model parameters of a lumped parameter physical model are identified by means of real-time collected forging force, punch speed and punch displacement data, and the initial model parameters are taken as a process fingerprint of a current workpiece. Subsequently, a reference energy dissipation rate curve is generated by performing adjustment operation on a preset reference energy dissipation rate curve according to the process fingerprint;

[0023] b) In the entire forging process, model parameters of the lumped parameter physical model are updated in real time;

[0024] c) A model predictive control strategy is adopted, and based on the model parameters continuously updated in step b) and the reference energy dissipation rate curve generated in step a), a control instruction for controlling the punch speed is generated by solving a constrained optimization problem in a limited prediction time domain through rolling;

[0025] d) while performing step c), calculate the prediction error of the lumped parameter physical model in parallel, and monitor the trend of the prediction error; when the prediction error is monitored to exhibit a preset abnormal trend, tighten the maximum forging pressure constraint in the optimization problem.

[0026] The present application constructs a system and method capable of adaptive control according to the real-time physical characteristics of the forging process. It no longer relies on fixed process programs, but through online identification of real-time working conditions, adaptive adjustment of control targets, and the use of forward-looking optimization control strategies, it realizes closed-loop control of the forging process. At the same time, by monitoring the prediction error of the model, key events are predicted and control constraints are adjusted, improving the robustness of the control process. Thus, the present application can adapt to the differences in the initial state of different billets and disturbances in the processing process, thereby improving the forming quality and consistency of aviation precision forgings.

[0027] The present application provides an aviation precision forging adaptive control system and method. It has the following beneficial effects:

[0028] 1、The present application sets an online dynamic model identification module, which can obtain and update model parameters reflecting the physical characteristics of the current billet in real time, and use a model predictive controller to control the real-time updated model parameters. Further, the target curve adaptive adjuster adjusts the control target individually at the initial stage of processing, so that the entire control strategy can adapt to the initial state differences of each billet and disturbances in the processing process from beginning to end, thereby enabling different batches of forgings to experience more consistent thermodynamic processing history, which helps to improve the quality stability of the final product.

[0029] 2、The present application sets an event triggering and constraint dynamic modulation module, which uses the prediction error of the lumped parameter physical model. The event triggering and constraint dynamic modulation module analyzes the trend of the prediction error to predict physical events such as the end of die filling where pressure is prone to sudden change, and actively and dynamically tightens the operating constraints of the model predictive controller before such events occur. This forward-looking adjustment mechanism avoids the pressure overshoot caused by lagging response in traditional control methods, thereby effectively protecting the die and equipment and reducing the risk of related forming defects.

[0030] 3、The control target is raised from traditional kinematics or dynamics to energy dissipation rate which can comprehensively reflect the physical nature of the machining process. Meanwhile, the model identification, target adjustment, predictive control and constraint modulation in the system constitute information coordination. For example, the results of model identification are used for predictive control and adjustment of control target; and the error of model prediction is used for modulation of control constraint. This multi-level and information-coupled control architecture enables the system to perform a globally optimized control strategy that takes into account performance and safety, and goes beyond the scope of simple feedback control. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 A functional block diagram of the adaptive control system for aviation precision forging forging and pressing machining according to an embodiment of the present application;

[0032] Figure 2 A flowchart of the adaptive control method for aviation precision forging forging and pressing machining according to an embodiment of the present application.

[0033] Among them, 10, online dynamic model identification module; 20, target curve adaptive adjuster; 30, model predictive controller; 40, event trigger and constraint dynamic modulation module. DETAILED DESCRIPTION

[0034] In order to make the purpose, technical scheme and advantages of the present application clearer, the embodiments of the present application will be further described in detail below with reference to the drawings. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0035] Referring to the drawings Figure 1 The adaptive control system for aviation precision forging forging and pressing machining can be deployed in a computer device or a dedicated industrial controller, and is communicatively connected with a forging press through a data interface.

[0036] The adaptive control system for aviation precision forging forging and pressing machining according to the present application can be implemented as one or more software modules running on a controller. The controller can be an industrial personal computer (IPC), an advanced function module of a programmable logic controller (PLC), or a dedicated control unit configured with a digital signal processor (DSP). The controller includes a processor, a memory for storing program instructions and data, and an input / output interface for receiving sensor data from the forging press and sending control instructions to it.

[0037] The adaptive control system for aviation precision forging forging and pressing machining includes an online dynamic model identification module 10, a target curve adaptive adjuster 20, a model predictive controller 30, and an event trigger and constraint dynamic modulation module 40.

[0038] The input end of the online dynamic model identification module 10 is connected with sensors arranged on the forging press, for receiving real-time collected forging force , punch speed and punch displacement data. The output end of the online dynamic model identification module 10 is connected with the target curve adaptive adjuster 20 and the model predictive controller 30, for providing the real-time updated model parameter vector to them.

[0039] The input end of the target curve adaptive adjuster 20 is connected with the online dynamic model identification module 10, for receiving the model parameters in the initial stage of the forging process. The output end of the target curve adaptive adjuster 20 is connected with the model predictive controller 30, for providing the adjusted generated reference energy dissipation rate curve to it.

[0040] The input end of the model predictive controller 30 is respectively connected with the online dynamic model identification module 10, the target curve adaptive adjuster 20 and the event-triggered and constraint dynamic modulation module 40. The output end of the model predictive controller 30 is connected with the actuator of the forging press, for outputting the calculated punch speed control instruction .

[0041] The input end of the event-triggered and constraint dynamic modulation module 40 is connected with the online dynamic model identification module 10, for receiving the prediction error of the lumped parameter physical model. The output end of the event-triggered and constraint dynamic modulation module 40 is connected with the model predictive controller 30, for sending the modulation signal to it when the preset condition is met.

[0042] In this embodiment, the overall working principle of the adaptive control system for the aviation precision forging forging process is as follows:

[0043] The adaptive control system establishes and updates a lumped parameter physical model capable of representing the physical characteristics of the current forging process in real time through the online dynamic model identification module 10.

[0044] In the initial stage of forging, the target curve adaptive adjuster 20 adjusts a reference energy dissipation rate target by using the initial model parameters of the lumped parameter physical model, to generate a control target adapted to the current specific blank.

[0045] Subsequently, in the entire process, the model predictive controller 30 generates the punch speed control instruction meeting the physical constraints of the equipment by using the real-time updated lumped parameter physical model and the generated control target through prospective optimization calculation.

[0046] Meanwhile, the event trigger and constraint dynamic modulation module 40 monitors the prediction performance of the model in parallel, and when a critical physical event (such as the end of the die filling) is foreseen, actively adjusts the operating constraints of the model predictive controller 30 to ensure the smoothness and safety of the process.

[0047] Referring to the drawings Figure 1 The online dynamic model identification module 10 is used to establish and update in real time a lumped parameter physical model that can represent the dynamic characteristics of the forging process. The online dynamic model identification module 10 receives real-time forging force , punch velocity and punch displacement data from sensors as inputs, and periodically calculates and outputs an updated model parameter vector .

[0048] In one specific embodiment, the online dynamic model identification module 10 uses a lumped parameter physical model to describe the relationship between the forging force and the punch motion state. The mathematical expression of the lumped parameter physical model is as follows:

[0049] ;

[0050] where, is the instantaneous forging force; is the punch displacement; is the punch velocity; is the strain rate sensitivity index determined in advance according to the material properties, which is regarded as a fixed constant during a process. The model parameters to be identified in the lumped parameter physical model are: the equivalent stiffness coefficient and the equivalent viscous damping coefficient .

[0051] The equivalent stiffness coefficient mainly represents the stiffness change effect of the forging due to the geometric constraint enhancement when filling the die cavity. The equivalent viscous damping coefficient comprehensively represents the plastic deformation resistance of the material itself and the friction effect between the forging and the die. The equivalent stiffness coefficient and the equivalent viscous damping coefficient together constitute the model parameter vector , whose expression is:

[0052] ;

[0053] To achieve online identification, the online dynamic model identification module 10 uses a recursive least squares (RLS) algorithm with a forgetting factor. In a discretized adaptive control system, for each discrete time step , the above lumped parameter physical model is converted into a linear regression form. Define the discrete time step the measured value of , the regression vector is then composed of the motion state data of discrete time steps: ;

[0054] The online dynamic model identification module 10 updates the model parameter vector by performing the following recursive calculation steps:

[0055] Step 1, calculate the prediction error based on the model parameters of the last time :

[0056] ;

[0057] Step 2, calculate the gain vector of the recursive least squares algorithm :

[0058] ;

[0059] wherein is the covariance matrix of the recursive least squares algorithm, is a forgetting factor, whose value ranges from 0 to 1, used to adjust the degree of influence of historical data on the current model parameter estimation.

[0060] Step 3, update the model parameter vector according to the prediction error and the gain vector :

[0061] ;

[0062] Step 4, update the covariance matrix for use at the next discrete time step:

[0063] ;

[0064] By repeatedly performing the above recursive calculation at each control cycle, the online dynamic model identification module 10 can continuously output a real-time updated model parameter vector that reflects the current processing conditions.

[0065] Referring to the accompanying Figure 1 , the target curve adaptive adjuster 20 is used to generate a one-time personalized control target for the forging process according to the initial physical characteristics of each blank. The target curve adaptive adjuster 20 is enabled at the initial stage of the forging process.

[0066] The target curve adaptive regulator 20 receives a set of model parameter vectors identified in the initial stage from the online dynamic model identification module 10, and defines the set of model parameter vectors as the process fingerprint of the current workpiece to be processed . The process fingerprint quantitatively characterizes the comprehensive physical properties of the initial temperature, dimensional tolerance and lubrication state of the current workpiece to be processed.

[0067] The target curve adaptive regulator 20 internally pre-stores a reference energy dissipation rate curve obtained through offline finite element simulation or calibration experiments . The reference energy dissipation rate curve represents the variation of energy dissipation rate with time under ideal, nominal working conditions to complete a forging process.

[0068] After obtaining the process fingerprint , the target curve adaptive regulator 20 performs mathematical operations on the reference energy dissipation rate curve according to a preset adjustment function to generate a new reference energy dissipation rate curve specific to the current working step. The generation process of the reference energy dissipation rate curve can be described by the following expression:

[0069] ;

[0070] wherein, is an adjustment factor calculated based on the process fingerprint . The adjustment function aims to quantify the deviation of the physical properties reflected by the process fingerprint as a correction amount to the reference target. The adjustment function can have various specific implementation forms.

[0071] In a specific embodiment, the adjustment factor is calculated only in association with the initial equivalent viscous damping coefficient contained in the process fingerprint . The mathematical expression of the process fingerprint is as follows:

[0072] ;

[0073] wherein, is the initial equivalent viscous damping coefficient value extracted from the process fingerprint ; and is a nominal equivalent viscous damping coefficient value representing ideal working conditions, which is pre-stored.

[0074] In another embodiment, the calculation of the adjustment factor can take into account the process fingerprint initial equivalent stiffness coefficient and initial equivalent viscous damping coefficient The mathematical expression of the adjustment factor can be a linear combination:

[0075]

[0076] wherein, is the initial equivalent stiffness coefficient value extracted from the process fingerprint ; is the nominal equivalent stiffness coefficient value representing the ideal working condition; and are preset weight coefficients for adjusting the degree of stiffness and viscous influence, and .

[0077] In yet another embodiment, the adjustment function may also be realized in a non-analytical way. For example, a multi-dimensional look-up table from the process fingerprint to the adjustment factor may be established in advance through a large number of experiments or simulations. In actual operation, the target curve adaptive adjuster 20 determines the final adjustment factor by looking up the table and interpolating between adjacent data points in the table according to the identified .

[0078] After the adjustment is completed, the generated reference energy dissipation rate curve is transmitted to the model predictive controller 30 and serves as the tracking target of the model predictive controller 30 in the entire subsequent forging process.

[0079] Referring to the accompanying drawings, Figure 1 the model predictive controller 30 is the core execution unit of the adaptive control system. The model predictive controller 30 is configured to calculate an optimal punch speed control instruction at each discrete time step of the forging process by solving a constrained finite-time-domain optimization problem, and output the instruction to the actuator of the forging press.

[0080] The model predictive controller 30 performs a rolling optimization at each discrete time step . The process of the rolling optimization is based on an optimization problem solved within a limited future time window (i.e., a prediction horizon). The optimization problem is defined by an objective function, a prediction model, and a set of constraint conditions.

[0081] The objective function is used to quantify the pros and cons of the control performance. In the present embodiment, the specific mathematical expression of the objective function is as follows:

[0082] ​;

[0083] wherein, is the index of the current discrete time step; is the value of the objective function; is the length of the prediction horizon, is the length of the control horizon, and ; is the actual measured state at the current discrete time step ; is the predicted value of the energy dissipation rate at the future discrete time step; is the target value of the reference energy dissipation rate curve provided by the target curve adaptive tuner 20 at the future discrete time step; is the sequence of punch velocity increments to be optimized for the future discrete time steps starting from the current discrete time step ; and is a pre-defined, positive definite weight matrix, matrix is used to penalize the deviation between the predicted energy dissipation rate and the reference target value, and matrix is used to suppress the variation amplitude of the control commands to ensure the smoothness of the control process.

[0084] The prediction model is used to calculate the predicted value in the objective function. The model predictive controller 30 receives the latest model parameter vector provided by the online dynamic model identification module 10. Based on the actual measured state at the current discrete time step and a candidate sequence of control increments , the model predictive controller 30 predicts the state sequence for the future discrete time steps by performing a forward integration or iteration on the dynamic of the adaptive control system. Subsequently, the future sequence of forging force is calculated using the state sequence and the model parameter vector . Finally, the future predicted sequence of energy dissipation rate is calculated according to the physical definition:

[0085] ;

[0086] The constraints define the physical boundaries and process requirements that must be satisfied when solving the optimization problem. These constraints are applied to the entire prediction horizon and can include, for example:

[0087] Punch velocity constraint: ;

[0088] wherein, and These are the minimum and maximum speeds allowed by the actuator of the forging press, respectively.

[0089] Punch speed change rate constraint:

[0090] ;

[0091] The punch speed change rate constraint is used to limit the drastic degree of speed change.

[0092] Forging pressure constraint:

[0093] ;

[0094] in, This is the maximum forging pressure allowed by the equipment. This maximum forging pressure can be dynamically adjusted by the event-triggered and constraint dynamic modulation module 40.

[0095] Terminal displacement constraints:

[0096] ;

[0097] in, It is the preset forging end point position.

[0098] At each discrete time step In this embodiment, the numerical optimizer within the model predictive controller 30 solves the aforementioned optimization problem once. Since the prediction model is linear and the objective function is quadratic in this embodiment, the optimization problem is constructed as a quadratic programming (QP) problem. The numerical optimizer used to solve this QP problem in real time can employ mature algorithms in the art, specifically including active-set methods or interior-point methods, to obtain an optimal control increment sequence. .

[0099] According to the rolling time-domain control principle, only the first element of the control increment sequence is controlled. It was actually adopted. The final speed command output to the actuator is: .

[0100] In the next discrete time step The adaptive control system will collect new sensor data and repeat the entire optimization process described above. (See attached document.) Figure 1, the event-triggered and constraint dynamic modulation module 40 operates as a parallel monitoring and intervention unit. The event-triggered and constraint dynamic modulation module 40 is configured to assess the prediction accuracy of the model used by the online dynamic model identification module 10 in real time, and to predict the upcoming physical events that can cause the forging force to rise sharply according to the assessment results, and then to actively adjust the operating constraints of the model predictive controller 30 before the events occur.

[0101] The event-triggered and constraint dynamic modulation module 40 continuously calculates the prediction error of the lumped parameter physical model . This prediction error is defined as the difference between the actual forging force measured by the sensor and the model predicted forging force calculated by the event-triggered and constraint dynamic modulation module 40 using the latest model parameter vector and the real-time measured state :

[0102] ;

[0103] wherein the model predicted forging force is calculated by:

[0104] ;

[0105] wherein, and are the latest model parameters at the current time received from the online dynamic model identification module 10.

[0106] The event-triggered and constraint dynamic modulation module 40 determines whether to trigger the constraint modulation by analyzing the trend of the prediction error over time. In one embodiment, the analysis of the trend is achieved by accumulating the prediction error within a sliding time window of a fixed length . The triggering logic is defined by the following condition:

[0107] ;

[0108] wherein, is a pre-set positive triggering threshold. When the absolute value of the accumulated sum of the prediction error within the sliding time window exceeds this triggering threshold, it indicates that there is a persistent and one-way systematic deviation in the model prediction. Such deviation is interpreted as a precursor to critical physical events such as the complete filling of the die or the obstruction of material flow.

[0109] Once the above triggering conditions are met, the event triggering and constraint dynamic modulation module 40 outputs a modulation signal to the model prediction controller 30. The modulation signal then modulates the forging pressure constraint in the optimization problem. Perform dynamic tightening. There are several specific ways to implement this tightening operation.

[0110] In one embodiment, a fixed scaling factor is used. Constraint on the original maximum forging pressure Reduce:

[0111] ;

[0112] in, It is a preset constant that takes a value between 0 and 1. For example, it is 0.05.

[0113] In another, more refined embodiment, the tightening amount of the maximum forging pressure constraint is correlated with the magnitude of the prediction error monitored at triggering. Its mathematical expression can be:

[0114] ;

[0115] in, It is a preset gain coefficient used to map the sum of prediction errors to the amount of forging pressure reduction.

[0116] The new maximum forging force constraint is triggered and calculated. Subsequently, this new maximum forging pressure constraint will be sent to the model predictive controller 30 to replace its original internal maximum forging pressure constraint. This information will be used in all subsequent optimization problems until the forging process is completed.

[0117] The following will refer to the appendix. Figure 2 This paper elaborates on embodiments of the adaptive control method for forging and pressing of aerospace precision forgings provided by the present invention. (Appendix) Figure 2 This is a flowchart of an adaptive control method for forging precision aerospace forgings according to an embodiment of the present invention. The method is executed on a forging machine configured with the adaptive control system of the present invention, and includes the following steps:

[0118] Step S100: System initialization.

[0119] Before a forging task begins, a system initialization process is performed first. This process includes loading and setting a series of preset parameters, including the strain rate sensitivity index in the lumped-parameter physical model. The prediction time domain length in the objective function of the model predictive controller 30 Control time domain length , weight matrix and ; forgetting factor of recursive least square algorithm used by online dynamic model identification module 10 ; and fixed length of sliding time window of event-triggered and constraint dynamic modulation module 40 and triggering threshold . Meanwhile, the reference energy dissipation rate curve and various types of original constraint values (maximum forging force and forging end position ) are loaded into the system memory.

[0120] Step S200: adaptive adjustment in initial stage of processing.

[0121] When the punch of the forging machine starts to descend and contacts the blank, the adaptive control system collects the forging force, punch speed and punch displacement data in real time. The online dynamic model identification module 10 continuously executes the recursive least square algorithm in the preset initial stage of the forging stroke (for example, the first 5% of the stroke) to identify the model parameter vector representing the initial physical characteristics of the blank, and solidifies the model parameter vector as the process fingerprint . Subsequently, the target curve adaptive adjuster 20 is triggered, which performs mathematical operations on the reference energy dissipation rate curve according to the obtained process fingerprint and the preset adjustment function (such as one of the aforementioned embodiments), to generate a reference energy dissipation rate curve adapted to the current blank. This adaptive adjustment step in the initial stage of processing is only executed once in a complete forging process.

[0122] Step S300: rolling execution of main control loop.

[0123] After completing step S200, the adaptive control system immediately enters the main control loop stage, which is repeatedly executed at each discrete time step until the processing end condition is met. At each discrete time step , the following operations are performed in sequence:

[0124] First, the online dynamic model identification module 10 updates the model parameter vector inside the online dynamic model identification module 10 from to by performing a recursive calculation of the recursive least square algorithm using the latest collected sensor data.

[0125] Then, the model predictive controller 30 receives this updated model parameter vector and based on the model parameter vector , the model parameter vector generated in step S200, and the current operating constraints, solve the constrained optimization problem inside it to calculate the optimal punch velocity increment sequence.

[0126] Finally, according to the principle of receding horizon control, only the first element of the punch velocity increment sequence is used to calculate the final punch velocity control command for the current discrete time step, and the final punch velocity control command is sent to the actuator of the forging press.

[0127] Step S400: Parallel event monitoring and constraint modulation.

[0128] This step S400 is executed in parallel with the main control loop in step S300 at every discrete time step . The event trigger and constraint dynamic modulation module 40 calculates the prediction error at every discrete time step and updates the cumulative sum of the prediction errors in its internal sliding time window. The event trigger and constraint dynamic modulation module 40 continuously compares the absolute value of the cumulative sum with a preset trigger threshold . If at a certain discrete time step , the absolute value of the cumulative sum exceeds , the event trigger and constraint dynamic modulation module 40 immediately calculates a new, tightened maximum forging pressure constraint and uses the new maximum forging pressure constraint to update and overwrite the original maximum forging pressure constraint stored in the model predictive controller 30.

[0129] Step S500: Process completion.

[0130] While executing steps S300 and S400, the adaptive control system continuously determines whether the current punch displacement has reached the preset forging end position . Once the condition is met, it is determined that the current forging process is complete, and the control process is terminated.

[0131] To further clarify the present application, the application process of the adaptive control system and method of the present application will be described below in conjunction with a specific embodiment. This embodiment describes the complete process of isothermal forging of a titanium alloy turbine disc forging using the technical solution of the present application.

[0132] In a specific application scenario, the adaptive control system of the present application is used to control a hydraulic press with a maximum forging pressure of 20000 kN.

[0133] First, system initialization in step S100 is performed before the process starts. The following parameters are set in the model predictive controller 30: strain rate sensitivity index of the material is set to 0.2; forgetting factor of the recursive least square algorithm is set to 0.98; prediction horizon of the model predictive controller 30 is 20 discrete time steps, control horizon is 5 discrete time steps; original maximum forging force constraint is set to 20000 kN; forging end position is set to 100 mm. At the same time, the benchmark energy dissipation rate curve for the turbine disk forging and the model parameters under nominal working condition, in which the nominal equivalent viscous damping coefficient is 50, have been pre-stored in the adaptive control system.

[0134] The forging starts, and the adaptive adjustment of the initial stage of the process in step S200 is entered. The punch descends and contacts the titanium alloy billet with an initial temperature slightly lower than the nominal temperature. In this stage, the punch displacement goes from 0 mm to 5 mm, the online dynamic model identification module 10 identifies the process fingerprint of the billet by real-time acquisition of the forging force, punch velocity and punch displacement data . Due to the lower temperature of the billet, its deformation resistance is higher, and the initial equivalent viscous damping coefficient identified is 55. The target curve adaptive adjuster 20 is then enabled, and calculates the adjustment factor according to and . The target curve adaptive adjuster 20 multiplies the adjustment factor by the benchmark energy dissipation rate curve , generates a new reference energy dissipation rate curve with an overall reduced amplitude, and transmits the new reference energy dissipation rate curve to the model predictive controller 30.

[0135] Subsequently, the control flow enters the main control loop in step S300 and the parallel event monitoring in step S400. During the movement of the punch displacement from 5 mm to 100 mm, the adaptive control system repeats this loop at every control period (every 10 milliseconds). The online dynamic model identification module 10 continuously updates the model parameters and to reflect the changes in the physical properties of the material due to work hardening and temperature changes. The model predictive controller 30 then uses the latest model parameter vectors and the generated reference energy dissipation rate curve , continuously rolling solution optimization problem, output optimal punch velocity control instruction to precisely track the energy dissipation rate target.

[0136] When the punch displacement reaches 92 mm, the forging material starts to fill the fine structure of the die in large quantities, causing the actual forging pressure to start a nonlinear growth faster than the model prediction. This makes the prediction error present a continuous positive deviation. The event-triggered and constraint dynamic modulation module 40 monitors the absolute value of the cumulative sum of the prediction error in its sliding time window with a length of 10 discrete time steps . The event-triggered and constraint dynamic modulation module 40 is triggered and immediately performs a constraint modulation operation. It calculates a new maximum forging pressure constraint according to a fixed scaling factor . The new maximum forging pressure constraint is sent to the model predictive controller 30, replacing its original 20000 kN constraint.

[0137] Finally, enter the processing end phase of step S500. In the final stroke of the punch displacement from 92 mm to 100 mm, the model predictive controller 30 must satisfy a more stringent upper limit of 19000 kN when performing optimization calculation on the predicted value of the forging pressure. This makes the model predictive controller 30 reduce the output punch velocity instruction in advance, thereby flattening the rising rate of the forging pressure when the die is completely filled, effectively avoiding pressure overshoot. When the punch displacement sensor measures , it is determined that the processing is complete, and the output control instruction is stopped, and the current forging process is ended.

[0138] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. An adaptive control system for forging and pressing of precision aerospace forgings, characterized in that, include: The online dynamic model identification module is used to collect forging pressure, punch speed and punch displacement data in real time during the forging process, and identify the model parameters of a lumped parameter physical model that describes the relationship between forging pressure and punch speed and punch displacement based on the data. The model prediction controller is used to receive the model parameters output by the online dynamic model identification module, and based on the model parameters and a preset reference energy dissipation rate curve, calculate and output control commands for controlling the speed of the punch by solving a constrained optimization problem. The event triggering and constraint dynamic modulation module is used to monitor the prediction error of the model used by the online dynamic model identification module in real time. When the prediction error is detected to show a preset abnormal trend, the module sends a modulation signal to the model prediction controller to adjust the constraints in the optimization problem. The event triggering and constraint dynamic modulation module is specifically configured as follows: by calculating the sum of the prediction error within the sliding time window, it is determined whether the prediction error shows a continuous unidirectional increasing trend, and when the trend meets the triggering condition, the modulation signal is sent to the model prediction controller to tighten the maximum forging pressure constraint in the optimization problem.

2. The adaptive control system for forging and pressing of aerospace precision forgings according to claim 1, characterized in that, The system also includes: The target curve adaptive adjuster is used to obtain the initial model parameters identified by the online dynamic model identification module as a process fingerprint in the initial stage of forging, and adjust a preset benchmark energy dissipation rate curve according to the process fingerprint to generate the reference energy dissipation rate curve used by the model prediction controller.

3. The adaptive control system for forging and pressing of aerospace precision forgings according to claim 2, characterized in that, The target curve adaptive adjuster is specifically configured as follows: The equivalent viscous damping coefficient in the initial model parameters is compared with the nominal equivalent viscous damping coefficient, and an adjustment factor is generated based on the comparison result. The adjustment factor is then used to scale the baseline energy dissipation rate curve to generate the reference energy dissipation rate curve.

4. The adaptive control system for forging and pressing of aerospace precision forgings according to claim 1, characterized in that, The lumped-parameter physical model used by the online dynamic model identification module has the following form: ； in, For forging pressure, For punch displacement, For the speed of the punch, The strain rate sensitivity index of the material, the model parameters include the equivalent stiffness coefficient. and equivalent viscous damping coefficient .

5. The adaptive control system for forging and pressing of aerospace precision forgings according to claim 1, characterized in that, The online dynamic model identification module uses recursive least squares with a forgetting factor to identify the model parameters online.

6. The adaptive control system for forging and pressing of aerospace precision forgings according to claim 1, characterized in that, The optimization problem solved by the model predictive controller includes the following objective function: ； in, This is the index of the current discrete time step; To be in discrete time steps Calculate the objective function value; To predict the length of the time domain, To control the length of the time domain; To be in discrete time steps For the future Predicted energy dissipation rate for each discrete time step; The reference energy dissipation rate curve is used in the future... The target value for each discrete time step; To be in discrete time steps Calculated, for future use The increment of the punch speed to be optimized at each discrete time step; and the weight matrix used to penalize the tracking error of the energy dissipation rate. This is a weight matrix used to suppress the incremental change in the punch speed.

7. An adaptive control method for forging and pressing of precision aerospace forgings, based on the system described in any one of claims 1-6, characterized in that, Includes the following steps: a) In the initial stage of forging, the initial model parameters of a lumped parameter physical model are identified in real time as a process fingerprint, and a reference energy dissipation rate curve is adjusted according to the process fingerprint to generate a reference energy dissipation rate curve. b) Update the model parameters of the lumped parameter physical model in real time throughout the entire forging process; c) Using a model predictive control strategy, based on the continuously updated model parameters in step b) and the reference energy dissipation rate curve generated in step a), control commands for controlling the speed of the punch are generated by rolling the solution of a constrained optimization problem. d) While performing step c), monitor the prediction error of the model in parallel. When the prediction error is detected to exhibit a preset abnormal trend, dynamically tighten the maximum forging pressure constraint in the optimization problem.

Images