Intelligent control method and system for wind power bearing machining

Abstract

The invention relates to the technical field of manufacturing and industrial process control, in particular to an intelligent control method and system for wind power bearing machining. The method comprises the following steps: acquiring monitoring data in a laser cladding process; based on the temperature dynamic prediction model, combined with the monitoring data, determining the predicted temperature of the next moment; according to the deviation between the predicted temperature and the target temperature, the adjustment amount of the laser power and the wire feeding speed is generated through optimization of a model prediction control strategy; the method also calculates a prediction residual error between the prediction temperature and the actual temperature at the next moment, and evaluates and adjusts model parameters of the dynamic temperature prediction model on line based on the prediction residual error to eliminate model mismatch; and finally, performing closed-loop control on a laser and a wire feeding system by applying the adjustment amount. According to the invention, through adaptive adjustment of the prediction model, the influence caused by sudden change of working conditions can be overcome, and the accuracy of temperature prediction is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of manufacturing and industrial process control, and particularly relates to an intelligent control method and system for wind power bearing processing. BACKGROUND

[0002] Copper-steel composite thrust bearing is applied to the fields of wind power and metallurgy due to its excellent wear resistance and mechanical strength. Laser cladding of the copper layer is a core process for realizing a high-performance composite structure, and the quality of the laser cladding directly determines the service life and operation reliability of the bearing.

[0003] At present, the traditional laser cladding method generally adopts fixed process parameters of laser power and wire feeding speed. However, in the actual production process, complex factors such as random fluctuations of the initial temperature of the steel substrate, nonlinear changes of the ambient temperature, and batch differences of material thermal properties can easily cause severe fluctuations of the temperature in the cladding area. Such temperature instability can directly lead to significant fluctuations of the metallurgical bonding strength of the copper layer, increase the probability of defects such as cracks, and finally cause a significant decline in the cladding quality, which is difficult to meet the stringent requirements of high-end equipment manufacturing industry on product consistency and reliability.

[0004] In related technologies, the idea of model-based predictive control is introduced, that is, a temperature prediction model is established, and the control quantity is adjusted in advance according to the deviation between the predicted future temperature and the target temperature. However, the temperature prediction model in these methods is usually established based on offline data calibration or theoretical derivation, and the model parameters are fixed. In the dynamic and complex laser cladding process, once the actual working condition or material property changes, the fixed prediction model will be mismatched with the actual physical process. Such model mismatch will lead to inaccurate temperature prediction, and then the control system makes wrong adjustments, which not only cannot realize accurate temperature control, but also may exacerbate temperature fluctuations and damage the stability of the control system. SUMMARY

[0005] To solve the technical problem of how to construct a temperature prediction model that can adapt to dynamic changes of working conditions and adjust model parameters when mismatched, and based on which to solve the adjustment amount of laser power and wire feeding speed to ensure the temperature control accuracy of laser cladding of copper-steel composite thrust bearing, the present application provides solutions in the following aspects.

[0006] In a first aspect, the present application provides an intelligent control method for wind power bearing processing, which comprises the following steps: Acquire monitoring data during the laser cladding process, the monitoring data including actual temperature, laser power and wire feeding speed; determine the predicted temperature at the next moment based on a preset temperature dynamic prediction model of the cladding zone and in combination with the monitoring data; optimize and solve the deviation between the predicted temperature and the preset target temperature, as well as the laser power adjustment amount and wire feeding speed adjustment amount to be solved, through a model prediction control strategy to generate the laser power adjustment amount and the wire feeding speed adjustment amount; calculate the prediction residual between the predicted temperature and the actual temperature at the next moment actually obtained; based on the prediction residual, evaluate the mismatch degree of the temperature dynamic prediction model; adjust the model parameters of the temperature dynamic prediction model according to the mismatch degree; apply the laser power adjustment amount and wire feeding speed adjustment amount to update the laser and wire feeding system to realize the control of the laser cladding processing process of the wind turbine bearing.

[0007] The present invention solves the core problem of model mismatch in the prior art, where the prediction model is static and cannot cope with changes in working conditions, by constructing a complete adaptive closed-loop control framework that includes acquisition-prediction-control-feedback correction. Compared with simple open-loop control or pure feedback control, this method achieves forward-looking prediction and regulation of temperature; compared with traditional model predictive control, this method introduces a mechanism for online adjustment of the model itself based on prediction residuals. This enables the control system to continuously learn and adapt to dynamic changes in the actual cladding process, thereby maintaining high-precision temperature control under complex working conditions, significantly improving the robustness, adaptability and ultimate stability of the cladding quality of the control system.

[0008] Preferably, the predicted temperature at the next moment satisfies the relationship: ; in, It is the cladding area The predicted temperature at the time; It is the cladding area The actual temperature at the moment; yes The influence coefficient of laser power on temperature at each moment; It is the cladding area Laser power at the moment; yes Time has come Laser power adjustment amount at the moment; yes The influence coefficient of wire feeding speed on temperature at each moment; It is the cladding area Wire feeding speed at all times; yes Time has come Wire feeding speed adjustment at each moment; It is the cladding area the time t the heat loss coefficient at the time t is a preset prediction time step.

[0009] The temperature dynamic prediction model is embodied as a linear mathematical expression with clear structure and explicit physical meaning. The model not only reasonably includes three key physical factors affecting temperature change, i.e. the heating effect of laser power, the cooling effect of wire feeding speed and the heat dissipation effect to the environment, but also makes the subsequent optimization and parameter identification efficient and easy to implement in calculation due to its linear form. This provides a solid foundation for the engineering application of the entire complex control algorithm, while ensuring the model description ability, effectively reducing the requirement for processor computing power, and enhancing the practicability and real-time performance of the method.

[0010] Preferably, the model parameters of the temperature dynamic prediction model include the influence coefficient of laser power on temperature, the influence coefficient of wire feeding speed on temperature and the heat loss coefficient.

[0011] Preferably, the optimization and solving by the model predictive control strategy include: constructing an objective function, and solving the laser power adjustment amount and the wire feeding speed adjustment amount by minimizing the objective function; the objective function satisfies the relationship: ; wherein, is the predicted temperature of the cladding zone at the time t is a preset target temperature of the cladding zone; is the predicted temperature of the cladding zone at the time t is the laser power adjustment amount from the time t is the laser power adjustment amount from the time t is the wire feeding speed adjustment amount from the time t is the wire feeding speed adjustment amount from the time t is the wire feeding speed adjustment amount from the time t is the wire feeding speed adjustment amount from the time t , are weight factors of the laser power adjustment amount and the wire feeding speed adjustment amount, respectively.

[0012] The application constructs and optimizes a specific objective function. The design of the objective function is extremely ingenious. It not only pursues the minimization of the deviation between the predicted temperature and the target temperature, but also suppresses the drastic changes of the laser power adjustment amount and the wire feeding speed adjustment amount by introducing weight factors. This multi-objective optimization strategy effectively avoids the frequent oscillation of the control system due to the pursuit of rapid response, makes the control process more smooth and stable, and is also conducive to prolonging the service life of the laser and other actuators.

[0013] Preferably, the solution to obtain the laser power adjustment amount and the wire feeding speed adjustment amount includes: also applying preset control constraints to the laser power adjustment amount and the wire feeding speed adjustment amount; and solving the laser power adjustment amount and the wire feeding speed adjustment amount to satisfy their corresponding control constraints.

[0014] Preferably, the evaluation of the mismatch degree of the temperature dynamic prediction model based on the prediction residual includes: The time window of the current moment is used to calculate the average absolute value of the prediction residual in the time window to obtain the prediction deviation; the mismatch degree of the temperature dynamic prediction model is the normalized prediction deviation; wherein the prediction deviation of the current moment is Satisfies the relationship: ;in, The first The actual temperature at the moment, The first The predicted temperature at the time, is the size of the time window at the current moment, is the absolute value symbol.

[0015] By calculating the mean absolute value of the prediction residuals within a preset time window, this method effectively smooths out the random interference caused by single-point measurement noise. Compared to using instantaneous residuals, the resulting prediction deviation more accurately and stably reflects the overall performance deviation of the prediction model over a recent period of time. This provides a more reliable and robust decision-making basis for subsequent model parameter adjustments, avoiding the adaptive mechanism's oversensitivity to noise and frequent misjudgments.

[0016] Preferably, adjusting the model parameters of the temperature dynamic prediction model according to the degree of mismatch includes: determining an adaptive forgetting factor according to the degree of mismatch, wherein the adaptive forgetting factor is negatively correlated with the degree of mismatch; and updating the model parameters of the temperature dynamic prediction model according to the prediction residual using a recursive least squares algorithm with the adaptive forgetting factor.

[0017] The present invention negatively correlates an adaptive forgetting factor with the degree of model mismatch. This design perfectly resolves the inherent conflict between response speed and noise immunity faced by traditional recursive least squares algorithms using a fixed forgetting factor. When the system undergoes sudden changes or the degree of mismatch is high, the forgetting factor automatically decreases, enabling the algorithm to quickly track changes. When the system is stable and the degree of mismatch is low, the forgetting factor automatically increases, enhancing noise suppression. This intelligent dynamic balancing mechanism ensures both agility and stability in the model parameter update process, which is key to improving overall system performance.

[0018] Preferably, the updating of the model parameters of the temperature dynamic prediction model includes: determining a correction term based on the product of the Kalman gain and the prediction residual; adding the model parameters at the previous moment to the correction term to obtain the updated model parameters at the current moment.

[0019] Preferably, the initial values ​​of the model parameters of the temperature dynamic prediction model are obtained by collecting multiple sets of sample data in the historical laser cladding process and solving the multiple sets of sample data using the least squares method.

[0020] In the second aspect, the present invention provides an intelligent control system for wind power bearing processing. The intelligent control system for wind power bearing processing includes a memory and a processor. The memory stores computer program instructions. When the computer program instructions are executed by the processor, an intelligent control method for wind power bearing processing according to the first aspect of the present invention is implemented.

[0021] By adopting the above technical solution, the intelligent control method for wind power bearing processing in the first aspect of the present invention is generated into a computer program and stored in a memory to be loaded and executed by a processor, thereby making a terminal device based on the memory and the processor for easy use.

[0022] The beneficial effects of the present invention are as follows: The present invention establishes a complete model adaptive closed loop: by calculating the prediction residual, assessing the degree of mismatch, and adjusting the model parameters, the prediction model can perform online self-correction based on the real-time prediction error. This design eliminates the dependence of the control system on a perfect, unchanging model, enabling it to continuously learn and adapt to the dynamic changes and uncertainties in the process, thereby demonstrating strong robustness and environmental adaptability in complex and ever-changing industrial sites. The present invention achieves intelligent regulation by making the adaptive forgetting factor negatively correlated with the degree of model mismatch: when the degree of mismatch is high, the forgetting factor automatically decreases, allowing the algorithm to focus more on current data, thereby rapidly updating the model parameters and achieving agile tracking of dynamic changes; when the degree of mismatch is low, the forgetting factor automatically increases, allowing the algorithm to focus more on historical data, effectively suppressing the interference of measurement noise on the model and ensuring smooth control. The objective function defined in the present invention minimizes the amplitude of change in the controlled variable while minimizing temperature deviation. This effectively avoids violent oscillations in the control output and makes the control process smoother. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A flow chart of an intelligent control method for wind power bearing processing provided by an embodiment of the present invention; Figure 2 This is a structural block diagram of an intelligent control system for wind turbine bearing processing provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The first aspect of the embodiment of the present application provides an intelligent control method for wind power bearing machining, as shown in the figure, the method comprises steps S100-S600: Figure 1 Step S100, acquiring monitoring data in the laser cladding process, the monitoring data comprising actual temperature, laser power and wire feeding speed.

[0025] It should be noted that in the copper-steel composite thrust bearing machining process, the copper layer laser cladding is the key process to ensure the mechanical strength of the product. The core of this process is to melt the copper wire by laser and form metallurgical bonding with the steel matrix. This step is the basis for closed-loop control. Real-time and accurate acquisition of core state variables such as temperature of the cladding area and current output of the control system such as laser power and wire feeding speed is the premise for subsequent state prediction, optimization solution and model correction. Only feedback control based on the true state of the current process can effectively suppress the impact of various disturbances on the cladding quality.

[0026] Specifically, a temperature measuring device such as an infrared thermometer or a two-color pyrometer is deployed near the laser cladding head to measure the center temperature of the cladding molten pool in real time, which is denoted as the actual temperature. At the same time, the laser power and wire feeding speed are acquired from the laser controller and wire feeding system. The laser cladding head is referred to as the cladding area below.

[0027] At this point, the actual temperature, laser power and wire feeding speed in the laser cladding process are acquired.

[0028] Step S200, determining the predicted temperature at the next time based on the preset temperature dynamic prediction model of the cladding area and in combination with the monitoring data.

[0029] It should be noted that the heat transfer in the laser cladding process is a complex dynamic process, and traditional control methods cannot cope with disturbances such as initial temperature of the steel matrix, environmental temperature and material batch differences, resulting in loss of temperature control. Therefore, building a dynamic model that can accurately predict the temperature of the cladding area is the premise and basis for realizing closed-loop intelligent control.

[0030] Specifically, based on the principle of energy conservation, a discrete-time dynamic recursive model describing the temperature change of the cladding area is established. The model relates the predicted temperature at the next time to the actual temperature at the current time, laser power, wire feeding speed and environmental heat dissipation.

[0031] According to the above construction logic, the predicted temperature at the next time predicted by the temperature prediction model satisfies the following relationship: ; Where, is the cladding area​ The predicted temperature at the time; It is the cladding area The actual temperature at the moment; yes The influence coefficient of laser power on temperature at the moment, in units of ; It is the cladding area The laser power at the moment, in units of ; yes Time has come The laser power adjustment amount at the moment, the unit is ; yes The influence coefficient of wire feeding speed on temperature at any moment, unit is ; It is the cladding area The wire feeding speed at the moment, in units of ; yes Time has come Wire feeding speed adjustment at each moment ; It is the cladding area Time has come The heat loss coefficient at the moment, the unit is ; is the preset prediction time step, in units of .

[0032] In this relationship, Represents Time, the total laser power The temperature change caused by the contributed energy input. The greater the laser power, the greater the value and the higher the temperature prediction value. Represents The temperature drops during this time due to the heat absorbed by the melted copper wire. The faster the wire feeding speed, the more heat is absorbed and the lower the temperature prediction value. Following the Newtonian cooling law, it describes The temperature drop is caused by natural heat dissipation driven by the temperature difference between the molten pool and the environment within a certain period of time. The higher the actual temperature at the current moment, the greater the temperature difference, the faster the heat dissipation, and the lower the temperature prediction value.

[0033] It should be noted that 、 Both are parameters that need to be controlled during the laser cladding process. How to make the values ​​of these two parameters more reasonable will be described in detail in the subsequent step S300.

[0034] It should also be noted that the model parameters 、 、 It changes dynamically over time. Its initial value 、 、 This can be achieved by collecting multiple sets of input laser power, wire feed speed, and output temperature data from historical cladding processes and fitting them using the offline least squares method. To cope with more complex changes during online operation, the present invention introduces a recursive least squares method with a forgetting factor in the subsequent step S400 to perform online adaptive updating of these parameters.

[0035] At this point, the predicted temperature at the next moment is obtained based on the preset temperature dynamic prediction model of the cladding zone.

[0036] Step S300: Based on the deviation between the predicted temperature and the preset target temperature, as well as the laser power adjustment amount and the wire feeding speed adjustment amount to be solved, an optimization solution is performed through a model predictive control strategy to generate the laser power adjustment amount and the wire feeding speed adjustment amount.

[0037] It should be noted that after obtaining the temperature dynamic prediction model, this embodiment further constructs a temperature control optimizer based on the model predictive control strategy to achieve precise and stable control of the cladding temperature. MPC is an advanced control strategy that uses models to predict future system behavior and determines the current optimal control input by solving an optimization problem. This allows for proactive handling of time delays, constraints, and multi-objective optimization problems.

[0038] Specifically, in each control cycle Time has come At this moment, the optimizer takes minimizing the deviation between the predicted temperature and the target temperature as its primary goal, while taking into account the adjustment range of the control input to avoid drastic fluctuations in the system.

[0039] According to the above logic, the objective function constructed Satisfies the relationship: ; in, It is the cladding area The predicted temperature at the time; is the preset target temperature of the cladding zone; yes Time has come Laser power adjustment amount at the moment; yes Time has come Wire feeding speed adjustment at each moment; 、 are the weight factors of laser power adjustment and wire feeding speed adjustment respectively.

[0040] In the equation, is the square of the temperature prediction error, which aims to make the predicted temperature as close as possible to the preset target temperature. and are the square penalty terms for controlling the laser power adjustment and wire feed speed adjustment, respectively, which are used to suppress the drastic changes in laser power and wire feed speed, thereby ensuring the smoothness of the control process and prolonging the life of the actuator.

[0041] The values of the weight factors and can be adjusted according to process requirements. For example, when laser cladding repair is performed on high-precision workpieces, if the temperature error is within the range of 80 , the values of and can be appropriately reduced, such as and , so that the controller quickly increases the laser power to pull the temperature back to the vicinity of the target value within 1-2 time steps, avoiding defects caused by excessively low temperature. However, for large-scale component laser surface quenching, it is necessary to ensure slow temperature rise and fall to avoid deformation or cracking of the component due to excessive thermal stress. If the temperature error is within the range of 20 , the weights are set as follows: and , so that the controller focuses more on limiting the drastic changes in power and smoothly reaches the target temperature within multiple time steps, avoiding thermal shock caused by sudden power changes.

[0042] It should be noted that in order to ensure that the control action is within the allowed range of the physical device and meets the process safety specifications, constraints must be imposed on the control adjustment.

[0043] Specifically, the control constraints are as follows: ; wherein is the laser power adjustment from time to time ; is the wire feed speed adjustment from time to time .

[0044] Based on the objective function and the constraints, the minimum and can be obtained. This optimization problem is a typical quadratic programming problem, which can be efficiently solved at each control period by using mature numerical optimization algorithms. The specific solution method is not described here.

[0045] Thus, through the optimization solution of model predictive control, the laser power adjustment and wire feed speed adjustment for the next control period are generated.

[0046] Step S400, calculating the prediction residual between the predicted temperature and the actual temperature actually obtained at the next moment; based on the prediction residual, evaluating the mismatch degree of the temperature dynamic prediction model; and adjusting the model parameters of the temperature dynamic prediction model according to the mismatch degree.

[0047] It should be noted that this step is the key to achieving adaptive control and improving robustness in the present invention. The laser cladding process is affected by unmodeled dynamics and disturbances such as material batch differences, changes in local heat dissipation conditions of the workpiece, and nozzle wear. This can cause the preset temperature dynamic prediction model to gradually mismatch and reduce prediction accuracy, thereby affecting the control effect. By comparing the predicted value with the actual value online and using their deviation to continuously correct the model parameters, the prediction model can always maintain the best approximation to the actual process, thereby ensuring that the control system can still maintain high performance in a changing environment.

[0048] Specifically, this embodiment defines a short-term prediction deviation , based on the current moment The recent performance of the model is measured by calculating the average absolute error between the model prediction value and the actual measurement value in the most recent time window. Satisfies the relationship: ; in, The first The actual temperature at the moment, The first The predicted temperature at the time, is the size of the time window at the current moment, is the absolute value symbol.

[0049] As a preferred embodiment, the window size The value of is a trade-off. , such as 10, can respond to changes faster but is more sensitive to measurement noise; larger For example, 20 is more robust to noise, but has a slower response. In this embodiment, it is preferred to To balance response speed and stability.

[0050] It should be noted that The numerical range of is not fixed, which is not convenient for direct use in subsequent logical judgment. Normalize it so that its value is The normalization process can be implemented using the Sigmoid function. The Sigmoid function is a prior art and will not be described in detail here.

[0051] At this point, the mismatch degree of the temperature dynamic prediction model is obtained.

[0052] Step S500: adjusting the model parameters of the temperature dynamic prediction model according to the mismatch degree.

[0053] It should be noted that if the model parameters in step S200 、 、 If the model remains unchanged, it will be difficult to adapt to dynamic fluctuations caused by factors such as raw material batches, surface treatment methods, and changes in environmental conditions. It will also be unable to accurately reflect the system's actual temperature response characteristics, resulting in a decrease in temperature prediction accuracy and poor control effectiveness. This invention uses a recursive least squares algorithm with a variable forgetting factor (VFF-RLS) to dynamically adjust the weights of historical data, giving more influence to recent data. This allows for rapid tracking of time-varying system characteristics, enabling online dynamic updates of model parameters, and effectively improving the model's adaptability to complex operating conditions.

[0054] Specifically, it includes steps S510 to S520: Step S510: Calculate an adaptive forgetting factor based on the mismatch degree.

[0055] It should be noted that the mismatch degree identified in step S400 needs to be converted into a direct control of the model parameter update rate. The adaptive forgetting factor adjustment mechanism in this embodiment is the bridge connecting the mismatch degree and the model parameter update rate.

[0056] According to the above logic, The adaptive forgetting factor at each moment satisfies the relationship: ; in, is The adaptive forgetting factor calculated at each moment for the next VFF-RLS update, is the normalized prediction deviation, 、 are the preset upper and lower limits of the forgetting factor respectively.

[0057] In this relationship, the forgetting factor The value of is negatively correlated with the normalized prediction deviation. When the system is stable, the normalized prediction deviation is ,but At this time, the VFF-RLS algorithm gives higher weight to historical data, and the model parameter updates slowly and steadily. At this time, the VFF-RLS algorithm quickly forgets the past data and gives higher weight to new data, so as to realize the rapid correction and re-convergence of the model parameters.

[0058] It should be pointed out that, in order to ensure the performance of the algorithm, should be close to 1 to ensure the stability of the model parameter estimation when the system is stable, and can be preferably 0.995. determines the maximum adaptive speed when the system mutates, and the smaller the value is, the faster the adaptation is, but instability can be introduced, and can be preferably 0.95.

[0059] Step S520, applying the VFF-RLS algorithm to update the model parameters with the adaptive forgetting factor.

[0060] In order to apply the VFF-RLS algorithm, first, the temperature prediction model in step S200 is rewritten into a standard linear regression form: ; Wherein, each variable is defined as follows: ; ; ; Wherein, represents the actual temperature change amount from to ; represents the regression vector constructed at ; is the actual temperature of the cladding zone at ; is the actual temperature of the cladding zone at ; is the laser power of the cladding zone at ; is the wire feeding speed of the cladding zone at ; is the preset prediction time step.

[0061] is the model parameter vector to be estimated. It is a time-varying vector, and its structure at any time is The goal of this algorithm is to calculate the optimal estimation value of based on new measurement data at each time. ​​Specifically refers to the current iteration step of the VFF-RLS algorithm, used The estimated values ​​of the model parameters known at time t.

[0062] The VFF-RLS algorithm is used to adjust the model parameter vector Perform real-time updates. The update process is as follows: First, calculate the Kalman gain , Satisfies the relationship: ; in, yes Kalman gain vector at time t; yes The covariance matrix of the time instant; is a column vector containing the All known input and state information that affects the system output at all times; It is the forgetting factor; is a column vector The transpose of is a row vector used for matrix and vector multiplication to comply with the operation rules of linear algebra.

[0063] This relationship is used to calculate The Kalman gain vector at time In the VFF-RLS algorithm, plays a crucial role as a weight, which determines how much we should trust the prediction error brought by the new measurement data when updating the model parameters. This means that the new data has a higher weight and the parameter adjustment will be larger; otherwise, it is more inclined to maintain the original parameter estimate.

[0064] Then, update the model parameter estimates, the model parameter estimates Satisfies the relationship: ; in, yes The model parameter vector at time t; yes The model parameter vector at time t; yes Kalman gain at time t; yes arrive The actual temperature change within a certain time period is also Output value at the moment; is a column vector The transpose of is a row vector used for matrix and vector multiplication to comply with the operation rules of linear algebra.

[0065] In this relationship, Is the prediction residual, the core driving force of the entire expression. It is the prediction made with the old parameters for the current input. The value in the square brackets represents the difference between the actual measured value and the model prediction value. If the prediction residual is zero, it means that the current model is perfect and the model parameters do not need to be updated. If the residual is not zero, the error signal will be passed through the gain To drive parameters Adjustments are made in the direction of reducing this error. This relationship is the core of the RLS algorithm, which performs the actual update of the model parameters. Its logic is that new estimate = old estimate + gain × prediction error, which is a classic predictor-corrector structure.

[0066] Finally, update the covariance matrix , the covariance matrix Satisfies the relationship: ; in, yes The covariance matrix of the time instant; It is the forgetting factor; yes The covariance matrix of the time instant; yes Kalman gain vector at time t; is a column vector The transpose of is a row vector used for matrix and vector multiplication to comply with the operation rules of linear algebra.

[0067] This relationship is used to update the covariance matrix for the next control cycle Prepare for calculations. yes The covariance matrix at the moment is the updated covariance matrix, which represents our estimated values ​​of the newly obtained model parameters The degree of uncertainty. Incorporating new measurement information at all times , we have a more certain understanding of the model parameters, so The value is usually higher than Small, indicating that the uncertainty is reduced. It will be used as the Kalman gain vector for the next cycle calculation. Input when .

[0068] At this point, the adjustment of the model parameters of the temperature dynamic prediction model is completed based on the mismatch degree.

[0069] Step S600: Apply the laser power adjustment amount and the wire feeding speed adjustment amount to update the laser and wire feeding system to achieve control of the laser cladding process of the wind turbine bearing.

[0070] In each control cycle arrive At this moment, by solving the above constrained optimization problem, that is, finding a set of objective functions Minimized and The system updates and outputs the control instructions for the next moment: ; ; in, It is the cladding area The laser power is updated at all times. It is the cladding area Laser power at the moment; yes Time has come Laser power adjustment amount at the moment; It is the cladding area Constantly updated wire feeding speed; It is the cladding area Wire feeding speed at all times; yes Time has come The wire feeding speed adjustment at each moment.

[0071] Updated control volume and The prediction-optimization-execution process is carried out in each control cycle, forming a complete closed-loop feedback control system.

[0072] It should also be noted that step S400 to step S500 form an adaptive closed loop. Assume that at a certain moment, the laser cladding head moves from the thick-walled main area of ​​the workpiece to a thinner edge position.

[0073] The local heat dissipation conditions of the workpiece undergo a physical mutation. Compared with the thick-walled body with huge heat capacity and good heat dissipation path, the heat is more likely to accumulate at the edge of the thin wall, and the heat dissipation rate is significantly slowed down. This means that the actual parameters describing the thermal response law of the system have changed, especially the coefficients related to heat dissipation. will suddenly decrease.

[0074] At this point, the model mismatch and error increase: the original model parameters suitable for thick wall areas It is no longer accurate at this new location. The prediction error term of the model, , its absolute value will increase significantly in the short term.

[0075] Adaptive response: According to the logic of step S400, a significant prediction error will immediately cause The normalized prediction deviation quickly approaches 1, which leads to the adaptive forgetting factor was quickly lowered to its lower limit .

[0076] Fast parameter correction: observing the Kalman gain The relationship in the denominator decrease, which will directly lead to the gain vector The absolute value of each component increases. In the update relation of The final effect is that the model parameters A major adjustment is made so that the model can quickly abandon the previously learned thermal response characteristics of the thick-walled area and quickly converge to parameter values ​​that can accurately describe the new characteristics of the current thin-walled edge.

[0077] At this point, the present invention has achieved precise and robust control of the temperature of the laser cladding zone under complex working conditions by constructing a complete control closed loop of prediction-optimization-evaluation-adaptation, significantly improving the consistency and processing stability of wind power composite bearing products.

[0078] The second aspect of this embodiment provides an intelligent control system for wind power bearing processing, such as Figure 2 As shown, the intelligent control system for wind power bearing processing includes a memory and a processor, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, an intelligent control method for wind power bearing processing according to the first aspect of the present invention is implemented.

[0079] The intelligent control system for wind turbine bearing processing also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

[0080] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as resistive random access memory, dynamic random access memory, static random access memory, enhanced dynamic random access memory, high bandwidth memory, hybrid memory cube, etc., or any other medium that can be used to store the required information and can be accessed by an application, module, or both. Any such computer storage medium may be part of, accessible to, or connectable to the device.

[0081] The above are all preferred embodiments of the present invention, and are not intended to limit the scope of protection of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. An intelligent control method for wind power bearing processing, characterized in that: Including steps: Acquiring monitoring data during the laser cladding process, wherein the monitoring data includes actual temperature, laser power, and wire feed speed; Based on a preset dynamic temperature prediction model of the cladding zone and in combination with the monitoring data, a predicted temperature at the next moment is determined; Based on the deviation between the predicted temperature and the preset target temperature, as well as the laser power adjustment amount and the wire feeding speed adjustment amount to be solved, an optimization solution is performed through a model predictive control strategy to generate the laser power adjustment amount and the wire feeding speed adjustment amount; Calculating a prediction residual between the predicted temperature and the actual temperature actually obtained at the next moment; and evaluating a mismatch degree of the temperature dynamic prediction model based on the prediction residual; Adjusting the model parameters of the temperature dynamic prediction model according to the degree of mismatch; The laser power adjustment amount and the wire feeding speed adjustment amount are applied to update the laser and the wire feeding system, thereby realizing control of the laser cladding processing process of the wind power bearing.

2. The intelligent control method for wind power bearing processing according to claim 1, characterized in that: The predicted temperature at the next moment satisfies the relationship: ； in, It is the cladding area The predicted temperature at the time; It is the cladding area The actual temperature at the moment; yes The influence coefficient of laser power on temperature at each moment; It is the cladding area Laser power at the moment; yes Time has come Laser power adjustment amount at the moment; yes The influence coefficient of wire feeding speed on temperature at each moment; It is the cladding area Wire feeding speed at all times; yes Time has come Wire feeding speed adjustment at each moment; It is the cladding area Time has come Heat loss coefficient at the moment; is the preset forecast time step.

3. The intelligent control method for wind power bearing processing according to claim 1, characterized in that: The model parameters of the temperature dynamic prediction model include the influence coefficient of laser power on temperature, the influence coefficient of wire feeding speed on temperature and the heat loss coefficient.

4. The intelligent control method for wind power bearing processing according to claim 1, characterized in that: The optimization solution using the model predictive control strategy includes: Constructing an objective function, and solving the laser power adjustment amount and the wire feeding speed adjustment amount by minimizing the objective function; The objective function Satisfies the relationship: ; in, It is the cladding area The predicted temperature at the time; is the preset target temperature of the cladding zone; yes Time has come Laser power adjustment amount at the moment; yes Time has come Wire feeding speed adjustment at each moment; 、 are the weight factors of laser power adjustment and wire feeding speed adjustment respectively.

5. The intelligent control method for wind power bearing processing according to claim 4, characterized in that: The solution to obtain the laser power adjustment amount and the wire feeding speed adjustment amount includes: Preset control constraints are also applied to the laser power adjustment amount and the wire feeding speed adjustment amount; The laser power adjustment amount and wire feeding speed adjustment amount are solved to meet their corresponding control constraints.

6. The intelligent control method for wind power bearing processing according to claim 1, characterized in that: The evaluating the mismatch degree of the temperature dynamic prediction model based on the prediction residual includes: Before the current moment The time window of the current moment is used as the time window, and the average absolute value of the prediction residuals in the time window is calculated to obtain the prediction deviation degree; The mismatch degree of the temperature dynamic prediction model is the normalized prediction deviation degree; Among them, the prediction deviation degree at the current moment Satisfies the relationship: ; in, The first The actual temperature at the moment, The first The predicted temperature at the time, is the size of the time window at the current moment, is the absolute value symbol.

7. The intelligent control method for wind power bearing processing according to claim 1, characterized in that: The adjusting the model parameters of the temperature dynamic prediction model according to the mismatch degree includes: determining an adaptive forgetting factor according to the mismatch degree, wherein the adaptive forgetting factor is negatively correlated with the mismatch degree; The model parameters of the temperature dynamic prediction model are updated according to the prediction residual using a recursive least squares algorithm with the adaptive forgetting factor.

8. The intelligent control method for wind power bearing processing according to claim 7, characterized in that: The updating of the model parameters of the temperature dynamic prediction model includes: determining a correction term based on a product of a Kalman gain and the prediction residual; The model parameters at the previous moment are added to the correction term to obtain the updated model parameters at the current moment.

9. The intelligent control method for wind power bearing processing according to claim 7, characterized in that: The initial values ​​of the model parameters of the temperature dynamic prediction model are obtained by collecting multiple sets of sample data in the historical laser cladding process and solving the multiple sets of sample data using the least squares method.

10. An intelligent control system for wind power bearing processing, characterized in that: The intelligent control system for wind power bearing processing includes a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, an intelligent control method for wind power bearing processing according to any one of claims 1 to 9 is implemented.

Images