A brake drum residual stress detection and shot blasting control parameter adaptive control method and system for an extreme temperature change environment

Abstract

The application provides a brake drum residual stress detection and shot blasting control parameter adaptive control method and system for an extreme temperature change environment, the method collects temperature, stress and strain data of the brake drum under extreme temperature change from room temperature to 600 DEG C through an embedded multi-source sensor array in real time; a finite element inversion algorithm of thermal-mechanical-phase change three-field coupling is used to dynamically reconstruct a three-dimensional residual stress field evolution model; based on the model, an optimal shot blasting parameter set is output through a multi-objective optimization decision model; a multi-nozzle collaborative control system is used to execute a differentiated shot blasting strategy, and adaptive processing is performed on the opening edge and the arc-shaped drum belly area of the brake drum; finally, quality parameters are detected online through a laser speckle interferometer and a microhardness tester, data is fed back to the decision model, a closed-loop control cycle is formed, and independent and continuous optimization of process parameters is realized, and the system correspondingly includes five modules of sensing detection, stress reconstruction, intelligent decision, execution control and quality feedback.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of intelligent manufacturing and precision detection, and particularly relates to a brake drum residual stress detection and shot blasting control parameter adaptive control method and system for extreme temperature change environment, and particularly relates to a technology for realizing accurate regulation and control of residual stress through multi-source sensing, digital twinning and closed-loop feedback. BACKGROUND

[0002] As a key safety component of the automobile braking system, the brake drum has an extremely harsh working environment, especially under the working condition of continuous downhill of heavy commercial vehicles, the surface temperature of the brake drum can rise from room temperature to above 600℃ in a short time, and then rapidly fall due to cooling, experiencing a severe extreme temperature change cycle. In this process, complex thermal-mechanical-phase change three-field coupling effects occur in the material, resulting in redistribution of residual stress, which significantly affects the fatigue life, dimensional stability and safety of the brake drum.

[0003] Currently, the detection of residual stress of the brake drum mainly adopts offline detection methods such as drilling method and X-ray diffraction. Although these methods have certain measurement accuracy, they have obvious limitations: the drilling method is a destructive detection method that will damage the integrity of the part and can only provide single-point static information; the X-ray diffraction method is expensive and has high environmental requirements, and cannot realize real-time online measurement. More importantly, the existing detection technology cannot capture the dynamic evolution process of residual stress under extreme temperature change environment, and cannot provide real-time data support for process control.

[0004] In terms of simulation analysis, existing researches, such as some simulation works based on ABAQUS, ANSYS and other software, mainly focus on thermal-structural coupling analysis or simulation of residual stress in specific manufacturing processes such as quenching process. Although these methods can provide certain theoretical guidance, they usually do not fully consider the real-time coupling effect among temperature field, phase change field and stress field, and the simulation results are heavily dependent on the setting of boundary conditions, making it difficult to be directly used for online control and optimization.

[0005] In terms of shot blasting strengthening process, current industrial practice generally adopts fixed parameter control method, which lacks adaptability to the actual stress state of the part and the temperature environment. This extensive control method has many drawbacks:

[0006] It cannot achieve differentiated processing, and cannot implement differentiated processing for different structural features of the brake drum, such as the opening edge and the arc-shaped drum belly area where stress concentration is serious, which can easily lead to insufficient processing or excessive strengthening;

[0007] It ignores environmental effects and fails to consider the influence of extreme temperature changes on the response characteristics of the material, such as the yield strength attenuation and stress relaxation effect of the material at high temperature, resulting in failure of the preset process parameters under actual working conditions.

[0008] Open-loop control, lack of feedback, existing systems lack effective online quality feedback mechanism, process parameter adjustment depends on manual experience, and closed-loop optimization control and autonomous decision-making in the production process cannot be realized.

[0009] Therefore, the prior art has insufficient sensing ability, lacks effective means for real-time, accurate and comprehensive sensing of the multi-physical field state of the brake drum, such as temperature, stress and strain, in an extreme temperature changing environment; in particular, existing sensing means, such as patch sensors, are difficult to work stably for a long time in harsh working conditions of high temperature, high speed and mechanical wear, and the high-temperature durability, installation reliability and signal integrity of the sensor are key technical bottlenecks restricting in-situ real-time monitoring; the model accuracy is limited, and there is a lack of a residual stress dynamic evolution model capable of accurately describing the thermal-mechanical-phase change multi-field coupling effect for guiding production; the control strategy is lagging, the existing shot blasting control strategy cannot be adaptively adjusted to adapt to changing working conditions and material states, the intelligent degree is low, the system lacks closed loop, the quality detection and process control links are disconnected, and an effective sensing-decision-execution-optimization closed-loop control mechanism cannot be formed. SUMMARY

[0010] In order to overcome the problems in the prior art that there is a lack of effective means for real-time, accurate and comprehensive sensing of the multi-physical field state of the brake drum in an extreme temperature changing environment, there is a lack of a residual stress dynamic evolution model capable of accurately describing the thermal-mechanical-phase change multi-field coupling effect, the existing shot blasting control strategy cannot be adaptively adjusted to adapt to changing working conditions and material states, and the quality detection and process control links are disconnected, and an effective closed-loop optimization mechanism cannot be formed, the present application discloses a brake drum residual stress detection and shot blasting control parameter adaptive control method and system for an extreme temperature changing environment, which can effectively solve the above technical problems.

[0011] To solve the above technical problems, the technical scheme of the present application is as follows:

[0012] A brake drum residual stress detection and shot blasting control parameter adaptive control method for an extreme temperature changing environment, comprising the following steps:

[0013] Through the multi-source sensor array arranged on the surface and inside of the brake drum, multi-sensor data fusion technology is used to collect and fuse the temperature gradient distribution, transient residual stress and micro-strain data of the brake drum in an extreme temperature changing environment in real time, wherein the extreme temperature changing environment refers to the working condition that the surface temperature of the brake drum rapidly cycles between room temperature and 600 DEG C;

[0014] A thermal-mechanical-phase change three-field coupled finite element inversion algorithm is used to dynamically reconstruct a three-dimensional residual stress field evolution model of the brake drum in an extreme temperature changing process based on the real-time data stream collected by the multi-source sensor array;

[0015] The three-dimensional residual stress field evolution model is input into a multi-objective optimization decision model trained by machine learning, which simultaneously considers the yield strength attenuation of the material at high temperature, the residual stress relaxation effect and the target compressive stress layer depth, and can receive subsequent quality feedback data to output an optimal shot blasting control parameter set adapted to the current temperature field and stress state;

[0016] According to the optimal shot blasting control parameter set, the shot blasting speed, projectile flow, jet angle, coverage time and nozzle motion trajectory of the shot blasting equipment are dynamically controlled through a multi-nozzle coordinated control system to realize differential shot blasting processing of the stress concentration areas of the arc-shaped structure and the opening edge of the brake drum.

[0017] After shot blasting processing, the surface strengthening layer quality parameters, including surface residual compressive stress, hardness and roughness, are detected in real time by an online microhardness tester and a laser speckle interferometer, and these detection data are fed back to the multi-objective optimization decision model as quality feedback data, and the model parameters are updated to form a closed-loop control cycle, realizing continuous optimization of the control parameters.

[0018] Preferably, the multi-source sensor array includes high-temperature resistant fiber optic grating sensors, pulse eddy current detection probes, and infrared thermal imaging instrument arrays, which work cooperatively in the following ways:

[0019] The high-temperature resistant fiber optic grating sensors are distributedly arranged and embedded and installed at a depth of 2-5 mm below the near surface of the brake drum by laser cladding technology, each sensor is spaced 30-50 mm in a grid pattern, directly measuring the micro-strain and temperature change of the brake drum under thermal-mechanical coupling load, and the working temperature range covers room temperature to 600℃;

[0020] The pulse eddy current detection probes are arranged in an array, with detection units uniformly distributed circumferentially around the brake drum, each detection unit containing an excitation coil and a detection coil, and through a multi-frequency scanning mode, residual stress distribution data within a depth of 5 mm below the surface of the brake drum are obtained, with a sampling interval not exceeding 0.5 seconds;

[0021] The infrared thermal imaging instrument array is composed of multiple high-speed infrared thermal imaging instruments, arranged in a multi-angle manner to achieve full coverage monitoring of the surface of the brake drum, with a temperature measurement accuracy of ±1℃, a spatial resolution not less than 0.5mm, a time resolution of 100Hz, and real-time acquisition of two-dimensional temperature field distribution data of the surface of the brake drum;

[0022] The multi-source sensor array collects data through a unified time synchronization signal, and the collected data is transmitted to a data processing center through a high-speed industrial Ethernet. A multi-sensor data fusion algorithm is used to perform space-time registration and fusion processing on the measurement data from different sources, generating a multi-physical field coupling data set containing temperature field, stress field and strain field, and providing complete input data for subsequent stress field reconstruction.

[0023] Preferably, the construction process of the thermal-mechanical-phase change three-field coupled finite element inversion algorithm includes:

[0024] The function relationship of the thermophysical property parameters of the brake drum material and the phase change kinetics model with the absolute temperature T is established, wherein:

[0025] The thermophysical property parameters include the thermal expansion coefficient α(T) and the thermal conductivity coefficient λ(T), and their function relationships are respectively:

[0026]

[0027]

[0028] wherein, and are specific coefficients determined by high-temperature performance experiments of the material;

[0029] The phase change kinetics model is a function f(t) describing the volume fraction of pearlite to austenite, which is described by the Avrami JMAK equation:

[0030]

[0031] wherein, the transformation rate constant k(T) itself is a function of the absolute temperature T, which is defined by the Arrhenius equation:

[0032]

[0033] In the formula, A is a pre-exponential factor, Q is a phase change activation energy, and R is a universal gas constant;

[0034] The multi-physical field coupling data set of the temperature field, the stress field and the strain field collected by the multi-source sensor array in real time is used as the boundary condition and the inversion input;

[0035] Through iterative calculation by the finite element model and combined with regularization processing, a three-dimensional residual stress field time-varying model is reconstructed.

[0036] Preferably, the multi-objective optimization decision model is obtained by training a deep reinforcement learning algorithm, and the training and decision process includes:

[0037] Define the state space, including real-time temperature field, stress field distribution data and material performance parameters reconstructed by the finite element inversion algorithm coupled by thermal-mechanical-phase change three fields;

[0038] Define the action space, corresponding to the combination of control parameters of shot intensity, jet angle, coverage time and processing path for the arc structure of the brake drum;

[0039] Set the reward function, with surface residual compressive stress value, hardness, roughness, and compressive stress layer depth and stress uniformity in the three-dimensional residual stress field evolution model as multi-objective optimization indicators;

[0040] The multi-objective optimization decision model dynamically adjusts the strategy according to the real-time state output, to control the shot blasting equipment to implement differentiated shot blasting processing on the stress concentration area of the arc structure and the opening edge of the brake drum.

[0041] Preferably, the differentiated shot blasting strategy includes:

[0042] For the stress concentration area of the opening edge of the brake drum, use higher shot blasting intensity and longer coverage time processing parameters;

[0043] For the arc drum abdomen area of the brake drum, use the optimized jet angle to ensure that the projectile flow vertically impacts the surface;

[0044] Implement the differentiated shot blasting strategy through the multi-nozzle cooperative control system.

[0045] Preferably, the closed-loop control cycle specifically includes:

[0046] Compare the result data detected online with the expected target value to generate a deviation signal;

[0047] Feedback the deviation signal to the multi-objective optimization decision model;

[0048] Use the optimization algorithm to update the parameters of the multi-objective optimization decision model online, to realize autonomous learning and continuous optimization of shot blasting process parameters.

[0049] Preferably, a brake drum residual stress detection and shot blasting control parameter adaptive control system for extreme temperature change environment, comprising:

[0050] Embedded multi-source sensing detection module for real-time acquisition of state data of the brake drum under extreme temperature change environment;

[0051] Stress field dynamic reconstruction and calculation module for running the thermal-mechanical-phase change three-field coupled finite element inversion algorithm;

[0052] Intelligent decision and optimization module, embedded with the multi-objective optimization decision model trained by machine learning;

[0053] A multivariate adaptive execution control module is used to receive instructions and drive the shot blasting equipment.

[0054] An online quality feedback and closed-loop control module is used for quality verification and parameter optimization.

[0055] Preferably, the high-temperature-resistant fiber grating sensor in the embedded multi-source sensing detection module is pre-installed at a specific depth position of the brake drum by laser cladding technology and can reliably work in an environment of 600 DEG C and above.

[0056] Preferably, the multivariate adaptive execution control module comprises:

[0057] A high-speed electromagnetic valve controlled projectile flow system is used to accurately control the projectile injection amount per unit time.

[0058] A six-degree-of-freedom robot arm is used to adjust the injection pose and angle according to program instructions.

[0059] An active temperature compensation unit is used to dynamically correct the relationship between the shot blasting energy and the material surface response according to the real-time temperature signal.

[0060] Preferably, the online quality feedback and closed-loop control module is integrated with a laser speckle interferometer and a microhardness tester, and can quickly detect the residual stress distribution, microhardness and surface morphology of the brake drum after shot blasting without interrupting the production rhythm.

[0061] The present application has the following advantages:

[0062] The present application realizes in-situ, real-time and multi-dimensional sensing in extreme environments, and solves the technical bottleneck that existing detection methods such as the drilling method and X-ray diffraction cannot perform real-time, full-range and non-destructive measurement in a high-temperature environment of 600 DEG C and a severe temperature change environment by fusing the high-temperature-resistant fiber grating sensor embedded by laser cladding, the pulse eddy current probe and the infrared thermal imager array, thereby providing a reliable data basis for precise control.

[0063] The modeling and reconstruction precision of residual stress dynamic evolution are improved, the finite element inversion model of thermal-mechanical-phase change three-field coupling is established, the regularization processing inverse problem is introduced, the errors caused by the traditional model ignoring the phase change effect or simplifying the coupling relationship are overcome, the three-dimensional residual stress field time-varying model of the brake drum in the extreme temperature change process can be reconstructed with high fidelity, and the dynamic evolution law is revealed.

[0064] The intelligentization, differentiation and adaptive control of the shot blasting process are realized, the multi-objective optimization decision model based on deep reinforcement learning can consider the high-temperature material performance attenuation, stress relaxation effect and complex structure characteristics simultaneously, output the optimal parameter set adaptive to the real-time state, and execute the differentiated strategy through the multi-nozzle cooperation system, completely change the extensive mode of the traditional fixed parameter shot blasting, and significantly improve the uniformity and consistency of the reinforcement effect;

[0065] A complete perception-decision-execution-optimization closed-loop control system is formed, through online quality detection and real-time data feedback, a closed-loop mechanism of process parameter autonomous learning and continuous optimization is established, so that the system has adaptive ability to cope with individual differences, environmental fluctuations and material performance attenuation, improves the intelligent level of the production process and the stability of the product quality, and finally, on the premise of ensuring uninterrupted production rhythm, significantly improves the fatigue life and use reliability of the brake drum. BRIEF DESCRIPTION OF DRAWINGS

[0066] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and those skilled in the art can also obtain other drawings according to the provided drawings without creative labor.

[0067] Figure 1 A step diagram of a brake drum residual stress detection and shot blasting control parameter adaptive control method for an extreme temperature change environment is provided for Embodiment 1 of the present application.

[0068] Figure 2 A structural block diagram of a brake drum residual stress detection and shot blasting control parameter adaptive control system for an extreme temperature change environment is provided for Embodiment 2 of the present application. DETAILED DESCRIPTION

[0069] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined application purposes, the exemplary embodiments will be described in detail, which are shown in the drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all the embodiments consistent with the present application. Instead, they are only examples of methods and systems consistent with some aspects of the present application as detailed in the appended claims.

[0070] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting. As used in this application and the appended claims, the singular forms "a," "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that the term "and / or," as used herein, refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0071] The specific embodiments according to the present application, features and effects thereof will be described in detail below with reference to the accompanying drawings and preferred embodiments.

[0072] Embodiment 1

[0073] Please refer to Figure 1 A brake drum residual stress detection and shot blasting control parameter adaptive control method for extreme temperature change environment includes the following steps:

[0074] Through the multi-source sensor array arranged on the surface and inside of the brake drum, the temperature gradient distribution, transient residual stress and micro-strain data of the brake drum under the extreme temperature change environment are collected and fused in real time by using multi-sensor data fusion technology, wherein the extreme temperature change environment refers to the working condition that the surface temperature of the brake drum rapidly cycles between room temperature and 600°C;

[0075] A thermal-mechanical-phase change three-field coupled finite element inversion algorithm is used to dynamically reconstruct the three-dimensional residual stress field evolution model of the brake drum during the extreme temperature change process based on the real-time data stream collected by the multi-source sensor array;

[0076] The three-dimensional residual stress field evolution model is input into the multi-objective optimization decision model trained by machine learning, which simultaneously considers the yield strength attenuation of the material at high temperature, the residual stress relaxation effect and the target compressive stress layer depth, and can receive subsequent quality feedback data to output the optimal shot blasting control parameter set adaptive to the current temperature field and stress state;

[0077] According to the optimal shot blasting control parameter set, the shot blasting speed, projectile flow, jet angle, coverage time and nozzle motion trajectory of the shot blasting equipment are dynamically controlled through the multi-nozzle cooperative control system to realize the differential shot blasting processing of the stress concentration areas of the arc-shaped structure and the opening edge of the brake drum;

[0078] After the shot blasting processing, the surface strengthening layer quality parameters including surface residual compressive stress, hardness and roughness are detected in real time by the online microhardness tester and laser speckle interferometer, and these detection data are fed back to the multi-objective optimization decision model as quality feedback data, and the closed-loop control cycle is formed by updating the model parameters to realize the continuous optimization of the control parameters.

[0079] Further, the multi-source sensor array includes high-temperature-resistant fiber grating sensors, pulse eddy current detection probes, and an infrared thermal imager array. The multi-source sensor array works cooperatively through the following manners:

[0080] The high-temperature-resistant fiber grating sensors are distributedly arranged and embedded and installed at a depth of 2-5 mm below the near surface of the brake drum through laser cladding technology. The depth range is in the maximum thermal gradient area during capturing the brake drum in operation, and at the same time, damage to the sensors caused by surface wear and oxidation is avoided. Each sensor is distributed in a grid shape with an interval of 30-50 mm. The interval is a critical spatial resolution determined according to the typical size of the brake drum, such as a diameter of 400-500 mm, the attenuation law of the stress field, and the finite element simulation, to ensure that the gradient change of the stress concentration area can be effectively captured, and key features are avoided to be missed due to too large interval or data redundancy due to too small interval. The sensors directly measure the micro-strain and temperature change of the brake drum under thermal-mechanical coupling load, adopt a special high-temperature coating for protection, and the working temperature range covers room temperature to 600℃. Each sensor node is equipped with an independent signal conditioning module for real-time temperature compensation and signal amplification processing.

[0081] The pulse eddy current detection probe is arranged in a ring array and is composed of 12 detection units. The number is determined based on the requirement that the circumferential resolution is not less than 30°. The detection probe is uniformly distributed around the circumference of the brake drum. Each detection unit includes an excitation coil and a detection coil. A high-frequency excitation mode is adopted, and the working frequency range is 200 kHz-8 MHz. Through a multi-frequency scanning mode, residual stress distribution data in a depth range of 5 mm below the surface of the brake drum can be obtained. The detection depth covers the main influence area of the shot peening layer, ensuring the correlation of the detection data and the process target. The sampling interval is not more than 0.5 seconds, the spatial resolution reaches 1 mm, the measurement accuracy is ±10 MPa, and the detection unit is equipped with an automatic calibration function to compensate the influence of temperature drift in real time.

[0082] The infrared thermal imager array is composed of 6 high-speed infrared thermal imagers. A multi-angle arrangement is adopted to realize full coverage monitoring of the surface of the brake drum. The working waveband of the thermal imager is 3-5 μm, the temperature measurement range is 0-800℃, the measurement accuracy reaches ±1℃, the thermal sensitivity is 0.03℃, the spatial resolution is not less than 0.5 mm, the time resolution reaches 100 Hz, and the automatic focusing and optical anti-shake functions are equipped to ensure that clear temperature field images can still be obtained in a vibrating environment.

[0083] The multi-source sensor array provides a unified time synchronization signal through a GPS synchronization clock, with a synchronization accuracy of 1 μs. The collected data is transmitted to the data processing center through a gigabit industrial Ethernet, with a transmission delay of less than 2 ms. A multi-sensor data fusion algorithm based on Kalman filtering is used to perform time and space registration and fusion processing on the measurement data from different sources. The data processing center first performs preprocessing on the original data, including noise filtering, outlier removal, and unit unification. Then, time and space registration is performed to establish a unified coordinate system and time reference. Finally, a weighted fusion algorithm is used to generate a multi-physical field coupling dataset containing temperature field, stress field, and strain field based on the measurement accuracy and reliability index of each sensor, providing complete input data for subsequent stress field reconstruction.

[0084] The entire sensor system is equipped with an environmental compensation module, which can monitor the environmental temperature, humidity, and vibration interference in real time and perform corresponding data compensation. The system also has a self-diagnosis function, which can monitor the working state of each sensor in real time. When an abnormal sensor is detected, the redundant sensor is automatically started or the measurement value is reconstructed to ensure the reliable operation of the system in extreme environments.

[0085] Further, the construction process of the heat-force-phase change three-field coupling finite element inversion algorithm includes:

[0086] The physical performance parameters include the thermal expansion coefficient α(T) and the thermal conductivity coefficient λ(T), and their functional relationships are respectively:

[0087]

[0088]

[0089] wherein, and are specific coefficients determined by high-temperature performance experiments of the material;

[0090] The phase change kinetics model is a function f(t) describing the volume fraction of pearlite to austenite transformation, which is described by the Avrami JMAK equation:

[0091]

[0092] wherein, the transformation rate constant k(T) itself is a function of absolute temperature T, which is defined by the Arrhenius equation:

[0093]

[0094] wherein, A is the pre-exponential factor, Q is the phase change activation energy, and R is the universal gas constant;

[0095] The finite element inversion algorithm of thermal-mechanical-phase transformation three-field coupling is a set of constitutive equations describing the material properties as a function of absolute temperature T. Each function is tightly coupled through the following physical mechanisms to form the calculation kernel of the inversion algorithm:

[0096] Thermal-mechanical coupling, thermal expansion coefficient a(T) connects the temperature field and the stress / strain field, indicating the linear expansion amount of the material caused by unit temperature rise, which is used to calculate the thermal strain; in finite element calculation, the temperature change of each element at the current time step directly calculates the thermal strain of the element, which will be input as the initial strain into the mechanical equilibrium equation to solve the thermal stress generated thereby, which is an important part of stress field calculation; coefficient Obtained by fitting the high-temperature expansion experiment of the brake drum material, usually HT250 gray cast iron, through a thermal mechanical analyzer;

[0097] Thermal-phase transformation-mechanical coupling, thermal conductivity coefficient l(T) follows Fourier's law and directly participates in the thermal conduction analysis and calculation of finite elements, which is used for thermal conduction analysis to calculate the internal temperature field distribution and evolution of the brake drum;

[0098] Phase transformation kinetics model f(t) quantifies the change rule of austenite transformation amount with time at a constant absolute temperature T. The phase transformation process is accompanied by volume change, such as the specific volume of austenite being smaller than that of pearlite, thereby generating phase transformation strain and further triggering the microstructure stress;

[0099] Transformation rate constant k(T) determines the dependence of phase transformation speed on temperature. The change of absolute temperature T directly affects f(t) through k(T), i.e. the temperature field drives the phase transformation process. The latent heat generated by phase transformation is fed back as internal heat source to affect the calculation of the temperature field, at the same time, the phase transformation strain is input as load to affect the stress field;

[0100] The temperature field propagates through l(T), generates thermal stress through a(T), and drives f(t) to describe the phase transformation process through k(T). The phase transformation process generates microstructure stress and reacts back to the temperature field;

[0101] A refined three-dimensional finite element grid model of the brake drum is established, and each element is assigned the above material function relationship;

[0102] The multi-physical field coupling data set of temperature field, stress field and strain field obtained by the multi-source sensor array after spatio-temporal registration is used as the input boundary condition and verification benchmark of the inversion algorithm;

[0103] Sequential coupling analysis method is used to solve the positive problem. First, the temperature field is obtained by solving the transient heat conduction equation; then the temperature field result is substituted into the phase transformation kinetics equation to calculate the phase transformation volume fraction and latent heat; finally, the temperature field and phase transformation strain are substituted into the mechanical constitutive equation to solve the stress field;

[0104] The regularization method is used to deal with the ill-posed inverse problem, and the internal parameters of the finite element model, such as the initial stress assumption, are iteratively adjusted to minimize the difference between the calculated results of the finite element model, such as the surface strain, and the measured data in the multi-source sensor array;

[0105] After iterative convergence, the algorithm finally outputs the time-varying model of the three-dimensional residual stress field with high confidence at the current time step, clearly showing the dynamic evolution process of internal stress in the entire cycle from room temperature to 600℃ and then cooling of the brake drum, including stress size, direction and distribution, especially the changes in stress concentration areas.

[0106] Further, the multi-objective optimization decision model is trained by a deep reinforcement learning algorithm, and the training and decision process includes:

[0107] The state perception module is used to receive and process the finite element inversion results from the finite element inversion algorithm of the thermal-mechanical-phase change three-field coupling, including temperature field distribution data, stress field distribution data and material performance parameters;

[0108] The decision generation module generates control decisions based on the current state information according to the deep neural network;

[0109] The reward calculation module calculates the reward value according to the control effect, which is used for model training and optimization;

[0110] The state space is defined, which contains the real-time temperature field, stress field distribution data and material performance parameters reconstructed by the finite element inversion algorithm of the thermal-mechanical-phase change three-field coupling;

[0111] The temperature field features include brake drum surface temperature distribution, temperature gradient distribution and temperature change rate; the stress field distribution data features include surface residual stress value, compressive stress layer depth and stress distribution uniformity; the material state performance parameter features include material phase change state and mechanical performance parameters;

[0112] The action space is defined, which corresponds to the combination of control parameters of the shot intensity, the jet angle, the coverage time and the processing path for the arc-shaped structure of the brake drum;

[0113] The shot intensity is realized by adjusting the pellet flow rate; the jet angle is realized by adjusting the nozzle attitude; the coverage time is realized by controlling the jet duration; and the processing path is realized by planning the nozzle motion trajectory;

[0114] The reward function is set, and the surface residual compressive stress value, hardness, roughness, compressive stress layer depth and stress uniformity in the three-dimensional residual stress field evolution model are used as multi-objective optimization indicators;

[0115] The reward function adopts a multi-objective weighted design method, the residual stress control accuracy measures the deviation of the actual residual stress value from the target value; the reinforcement layer quality evaluation considers the depth and uniformity of the compressive stress layer; the processing efficiency considers the process processing time factor; and the energy consumption index evaluates the energy consumption level;

[0116] The multi-objective optimization decision model outputs dynamic adjustment strategies according to real-time state to control the arc structure and opening edge stress concentration area of the brake drum of the shot blasting equipment to implement differentiated shot blasting processing;

[0117] The training process of the multi-objective optimization decision model includes the following steps:

[0118] Environment building, establishing a virtual training environment, integrating a finite element simulation model;

[0119] Data collection, collecting state-action-reward sample data through random exploration;

[0120] Network training, using a deep neural network to learn the optimal control strategy;

[0121] Strategy evaluation, verifying the effectiveness of the training results through simulation;

[0122] In actual application process, the multi-objective optimization decision model runs according to the following process:

[0123] Real-time acquisition of stress field reconstruction results provided by the finite element inversion algorithm of thermal-mechanical-phase change three-field coupling;

[0124] Extract state features and input the trained decision model;

[0125] Output the optimal control parameter combination;

[0126] Send the control parameters to the actuator;

[0127] Collect the processing results for online updating of the model;

[0128] Establish a regular updating mechanism:

[0129] Collect production data every week to update the training data set;

[0130] Fine-tune and optimize the model every month;

[0131] Comprehensive retraining is performed every quarter;

[0132] Establish a version management system to track model changes.

[0133] Further, the differentiated shot blasting strategy is based on the control parameters output by the multi-objective optimization decision model, and implements partition processing according to the structural characteristics of the brake drum, including:

[0134] According to the structural characteristics and stress distribution characteristics of the brake drum, the processing area is divided into an opening edge stress concentration area and an arc drum belly area;

[0135] For the opening edge stress concentration area of the brake drum, as a key stress part of the brake drum, higher shot blasting intensity and longer coverage time are adopted. By increasing the shot blasting intensity and extending the processing time, sufficient surface strengthening of the stress concentration area is ensured. In specific implementation, the system automatically identifies the position and geometric characteristics of the edge area and calls the preset strengthening processing parameter combination;

[0136] For the arc drum belly area of the brake drum, an optimized jet angle is adopted to ensure that the pellet flow vertically impacts the surface, focusing on controlling the jet angle to ensure that the pellet flow is always perpendicular to the curved surface, ensuring the uniformity of the processing effect. The system adjusts the nozzle posture in real time according to the curvature change of the curved surface to maintain the best impact angle;

[0137] The above-mentioned differential processing is realized by a multi-nozzle cooperative control system, which adopts a multi-nozzle cooperative working mode:

[0138] The system is equipped with two special nozzles, one for high-intensity processing of the edge area and the other for regular processing of the drum belly area. Each nozzle is equipped with an independent parameter adjustment device and a motion control system;

[0139] A time sequence alternation working mode is adopted to avoid mutual interference during processing. First, the edge processing nozzle completes the strengthening processing of the opening edge area, and then the drum belly processing nozzle performs regular processing of the arc curved surface area. The motion trajectory and working time of the two nozzles are coordinated by the central controller;

[0140] Based on the three-dimensional residual stress field evolution model of the brake drum, the optimal motion trajectory of each nozzle is pre-planned. In actual processing, the relative position and attitude of the nozzle and the workpiece surface are accurately controlled through real-time position detection and feedback to ensure accurate execution of the processing parameters;

[0141] In the process monitoring and adjustment, the shape and coverage of the pellet flow are observed in real time through a visual monitoring system, and the acoustic sensor is used to monitor the impact sound characteristics to indirectly evaluate the consistency of the processing intensity;

[0142] According to the monitoring results, the processing parameters are adjusted in real time. When the processing effect deviates from the expected result, the system automatically adjusts the relevant parameters to ensure that the final processing quality meets the requirements;

[0143] The processing parameters and process data of each workpiece are recorded in detail to establish a process database. These data are used for subsequent process optimization and quality traceability.

[0144] Further, the closed-loop control cycle specifically includes:

[0145] Data acquisition and preprocessing, receiving the surface hardness value measured by the online microhardness tester, the surface residual compressive stress value measured by the laser speckle interferometer, and the surface roughness value, filtering and denoising the collected raw data, and uniting the preprocessing to ensure data effectiveness;

[0146] Target value comparison, comparing the above preprocessed real-time detection data, surface hardness value, surface residual compressive stress value and surface roughness value with the expected target value set by the multi-objective optimization decision model one by one, and the expected target value is the optimal value range set in advance according to the product technical specifications and process requirements;

[0147] Comprehensive deviation calculation, a deviation signal representing the overall quality deviation degree is generated by using a weighted comprehensive deviation algorithm;

[0148] The deviation signal together with the corresponding real-time process parameter set and real-time state data, the real-time process parameter set is the surface hardness value measured by the online microhardness tester, the surface residual compressive stress value measured by the laser speckle interferometer, and the surface roughness value, and the real-time state data is the multi-physical field coupling data set of temperature field, stress field and strain field collected by the multi-source sensor array in real time, together form a training sample data, and is stored in the experience replay buffer of the multi-objective optimization decision model as the training data for model updating;

[0149] The system continuously monitors the deviation signal and the buffer data volume, and automatically triggers the incremental learning process of the model when the deviation continuously exceeds the threshold or the buffer data accumulates to a certain scale;

[0150] The parameters of the multi-objective optimization decision model are updated online by using the optimization algorithm to realize the autonomous learning and continuous optimization of the shot blasting process parameters;

[0151] The optimization algorithm uses optimization algorithms such as stochastic gradient descent (SGD) to update the internal parameters of the deep reinforcement learning model with the goal of minimizing the loss function composed of the deviation signal;

[0152] A small batch learning method is used to randomly select a batch of historical samples from the experience replay buffer, including successful and failed cases for each parameter update to improve training stability and efficiency;

[0153] By updating the parameters of the multi-objective optimization decision model, the multi-objective optimization decision model can output more accurate shot blasting parameters when facing similar stress field and temperature field states in the future, thereby reducing the deviation between predicted quality and actual quality, and realizing autonomous learning and continuous optimization;

[0154] The multi-objective optimization decision model is updated and verified in a simulation environment to evaluate the performance of the multi-objective optimization decision model under new parameters, and the online control can be put into operation through the rear;

[0155] The multi-objective optimization decision model of the historical version is reserved, and if the performance of the new model decreases, it can be quickly rolled back to ensure the stability of the production process.

[0156] Embodiment 2

[0157] Please refer to Figure 2 A brake drum residual stress detection and shot blasting control parameter adaptive control system for extreme temperature change environment, comprising:

[0158] An embedded multi-source sensing detection module for real-time acquisition of state data of the brake drum under extreme temperature change environment;

[0159] A stress field dynamic reconstruction and calculation module for running a thermal-mechanical-phase change three-field coupled finite element inversion algorithm;

[0160] An intelligent decision and optimization module embedded with a multi-objective optimization decision model trained by machine learning;

[0161] A multivariable adaptive execution control module for receiving instructions and driving the shot blasting equipment;

[0162] An online quality feedback and closed-loop control module for quality verification and parameter optimization.

[0163] Further, the high-temperature resistant fiber Bragg grating sensor in the embedded multi-source sensing detection module is pre-installed at a specific depth position of the brake drum by laser cladding technology and can reliably work in an environment of 600°C and above;

[0164] The high-temperature resistant fiber Bragg grating sensor is made of germanium-boron co-doped optical fiber that has been annealed at 700°C and is metalized and packaged, and the package body is made of inconel alloy pipe material and is sealed by argon arc welding, so that it has the basic performance of resisting 600°C high temperature;

[0165] The high-temperature resistant fiber Bragg grating sensor is pre-installed at a specific depth position of the brake drum by laser cladding technology and can reliably work in an environment of 600°C and above;

[0166] Pretreatment of the substrate, in the stress sensitive area of the brake drum to be embedded with the sensor, use pulse laser to clean out a positioning groove with a width of 1.0mm and a depth of 0.8mm;

[0167] Sensor positioning, accurately place the packaged sensor in the groove, and temporarily fix it with high-temperature ceramic glue;

[0168] Laser cladding, FeCrAl iron-based alloy powder with a particle size of 50-150 μm is selected as the cladding material, a semiconductor laser is used for cladding under argon protection, the energy input is controlled to ensure that the temperature peak of the molten pool does not exceed 750℃ and the duration is shorter than 0.5 seconds to avoid sensor failure due to overheating;

[0169] Post-processing, after the cladding is completed, the fusion area is finely ground and polished to make it flush with the surface of the brake drum, and the sensor is recalibrated.

[0170] Further, the multivariable adaptive execution control module comprises:

[0171] A high-speed electromagnetic valve control projectile flow system is used to accurately control the amount of projectile injection per unit time; the flow setting value (unit: kg / min) output by the multi-target optimization decision model is received, the required electromagnetic valve opening curve is calculated based on the setting value by the high-speed controller based on FPGA, the electromagnetic valve is driven by the PWM signal to accurately control the projectile flow, the actual flow value is fed back in real time by the flow sensor to form a closed-loop control, the control parameters are adjusted every 100 ms to ensure the flow control accuracy;

[0172] A six-degree-of-freedom robot arm is used to adjust the injection pose and angle according to program instructions; an optimal injection path is generated based on the brake drum CAD model to ensure that the nozzle axis is always at an angle of ≤10 with the normal direction of the workpiece surface, the moving speed is adjusted in real time according to the processing requirements, such as 50-200 mm / s, a torque sensor is built-in to monitor the collision risk in real time; a spiral asymptotic path is used in the opening edge area to ensure uniform coverage; an equidistant scanning path is used in the arc drum belly area to maintain a constant injection distance; a smooth interpolation algorithm is used in the transition area to avoid path mutation;

[0173] An active temperature compensation unit is used to dynamically correct the relationship between the shot blasting energy and the material surface response according to the real-time temperature signal; 8 temperature measurement points are arranged at key positions of the brake drum, the sampling frequency is 100 Hz, the temperature resolution is 0.1℃, wireless transmission is used to avoid line interference, a temperature-plasticity curve database is established for different materials, and the compensation parameters are updated every 50 ms; the projectile flow, injection angle and action time are adjusted according to the temperature;

[0174] The overall workflow of the multivariable adaptive execution control module is to receive the control instruction set of the system, the active temperature compensation unit first calculates the parameter correction amount, the six-degree-of-freedom robot arm moves to the target position and adjusts the posture, the high-speed electromagnetic valve is opened according to the corrected parameters to control the projectile flow, and the state information of each subsystem is exchanged in real time to ensure collaborative work, and the execution result is fed back to the upper system after the processing is completed.

[0175] Further, the online quality feedback and closed-loop control module integrates a laser speckle interferometer and a microhardness tester. The laser speckle interferometer and the microhardness tester adopt a parallel mechanical arm integrated architecture, and realize automatic switching of detection functions through a quick switching device, so that the residual stress distribution, microhardness and surface morphology of the shot-blasted brake drum can be quickly detected online without interrupting the production rhythm.

[0176] The laser speckle interferometric detection is implemented by using a dynamic laser speckle interferometric technology, which can realize stress measurement without surface treatment. A GPU-accelerated phase extraction algorithm is used to generate a residual stress distribution cloud map of the brake drum surface.

[0177] The microhardness detection is implemented by automatically positioning to the key area for hardness detection according to the stress detection result, using a new dynamic indentation method and integrating an automatic indentation recognition algorithm to realize unmanned operation.

[0178] Key quality indicators such as surface residual compressive stress value, stress distribution uniformity and hardness value are calculated in real time. The detection frequency and density are dynamically adjusted according to the quality condition. Real-time trend analysis is performed on the detection data to predict quality changes. The detection results are transmitted to a multi-target optimization decision model in real time. Process parameter adjustment suggestions are generated according to the detection results. The detection strategy and parameter adjustment rules are optimized based on historical detection data. The detection process is carried out in parallel with the production process to realize zero-waiting detection. When quality abnormalities are found, the adjustment mechanism is automatically triggered without affecting normal production.

Claims

1. A brake drum residual stress detection and shot blasting control parameter adaptive control method for extreme temperature change environment, characterized in that, The method comprises the following steps: Through the multi-source sensor array arranged in the embedded mode on the surface and inside of the brake drum, the temperature gradient distribution, transient residual stress and micro-strain data of the brake drum under extreme temperature change environment are collected and fused in real time by using multi-sensor data fusion technology, wherein the extreme temperature change environment refers to the working condition that the surface temperature of the brake drum rapidly cycles between room temperature and 600 DEG C; A three-field coupled finite element inversion algorithm of heat-force-phase change is used to dynamically reconstruct the three-dimensional residual stress field evolution model of the brake drum in the extreme temperature change process based on the real-time data stream collected by the multi-source sensor array, and the construction process of the three-field coupled finite element inversion algorithm of heat-force-phase change includes: The function relationship of the thermal physical performance parameters and the phase change kinetics model of the brake drum material with the absolute temperature T is established, wherein: The thermal physical performance parameters include the thermal expansion coefficient alpha (T) and the thermal conductivity coefficient lambda (T), and the function relationship is respectively: ； ； wherein , , , and , , are specific coefficients determined by high temperature performance experiments of the material; The phase change kinetics model is a function f (t) for describing the volume fraction of the pearlite to austenite transformation, which is described by the Avrami JMAK equation: ； Wherein, the transformation rate constant k (T) is a function of the absolute temperature T, which is defined by the Arrhenius equation: ； In the formula, A is a pre-factor, Q is a phase change activation energy, and R is a universal gas constant; The multi-physical field coupling data set of the temperature field, stress field and strain field collected by the multi-source sensor array in real time is taken as the boundary condition and inversion input; The three-dimensional residual stress field time-varying model is reconstructed through iterative calculation of the finite element model and combined with regularization processing; The three-dimensional residual stress field evolution model is input into the multi-objective optimization decision model trained by machine learning, which simultaneously considers the yield strength attenuation of the material at high temperature, the residual stress relaxation effect and the target compressive stress layer depth, and can receive subsequent quality feedback data to output the optimal shot blasting control parameter set adaptive to the current temperature field and stress state; According to the optimal shot blasting control parameter set, the shot blasting speed, pellet flow, jet angle, coverage time and nozzle motion trail of the shot blasting equipment are dynamically controlled through the multi-nozzle cooperative control system, so as to realize the differential shot blasting treatment of the stress concentration areas of the arc structure and the opening edge of the brake drum; After the shot blasting treatment, the surface strengthening layer quality parameters including the surface residual compressive stress, hardness and roughness are detected in real time by the online microhardness tester and laser speckle interferometer, and these detection data are taken as the quality feedback data and returned to the multi-objective optimization decision model, so that the closed-loop control cycle is formed by updating the model parameters to realize the continuous optimization of the control parameters.

2. The brake drum residual stress detection and shot blasting control parameter adaptive control method for extreme temperature change environment, according to claim 1, characterized in that, The multi-source sensor array comprises high-temperature resistant fiber grating sensors, pulse eddy current detection probes and infrared thermal imager arrays, and the multi-source sensor array cooperates in the following manner: The high-temperature resistant fiber grating sensors are arranged in a distributed manner and are embedded and installed at a depth of 2-5 mm below the near surface of the brake drum by laser cladding technology, each sensor is distributed in a grid shape with a spacing of 30-50 mm, directly measures the micro-strain and temperature change of the brake drum under thermal-force coupling load, and the working temperature range covers room temperature to 600 DEG C; The pulse eddy current detection probe is arranged in an array, and the detection units are uniformly distributed around the circumference of the brake drum. Each detection unit includes an excitation coil and a detection coil. The residual stress distribution data of the surface and the depth of 5 mm below the surface of the brake drum are obtained through multi-frequency scanning mode, and the sampling interval is not more than 0.5 seconds; The infrared thermal imager array is composed of multiple high-speed infrared thermal imagers, and is arranged in a multi-angle manner to realize full-coverage monitoring of the surface of the brake drum. The temperature measurement accuracy is ±1℃, the spatial resolution is not less than 0.5 mm, the time resolution is 100 Hz, and the two-dimensional temperature field distribution data of the surface of the brake drum are obtained in real time. The multi-source sensor array collects data through a unified time synchronization signal, and the collected data is transmitted to the data processing center through a high-speed industrial Ethernet. A multi-sensor data fusion algorithm is used to perform space-time registration and fusion processing on the measurement data from different sources, and a multi-physical field coupling data set including temperature field, stress field and strain field is generated, providing complete input data for subsequent stress field reconstruction.

3. The brake drum residual stress detection and shot blasting control parameter adaptive control method for extreme temperature change environment, according to claim 1, characterized in that, The multi-objective optimization decision model is trained by a deep reinforcement learning algorithm, and the training and decision process includes: Defining a state space including real-time temperature field, stress field distribution data and material performance parameters reconstructed by the thermal-mechanical-phase change three-field coupling finite element inversion algorithm; Defining an action space corresponding to the combination of control parameters such as shot intensity, jet angle, coverage time and processing path for the arc-shaped structure of the brake drum; Setting a reward function, and taking the surface residual compressive stress value, hardness, roughness, and stress uniformity in the three-dimensional residual stress field evolution model as multi-objective optimization indexes; The multi-objective optimization decision model outputs a dynamic adjustment strategy according to the real-time state to control the shot blasting equipment to implement differential shot blasting on the arc-shaped structure and the stress concentration area of the opening edge of the brake drum.

4. The brake drum residual stress detection and shot blasting control parameter adaptive control method for extreme temperature change environment, according to claim 1, characterized in that, The differential shot blasting strategy includes: For the stress concentration area of the opening edge of the brake drum, higher shot blasting intensity and longer coverage time are used; For the arc-shaped drum belly area of the brake drum, an optimized jet angle is used to ensure that the projectile flow vertically impacts the surface; The differential shot blasting strategy is realized through a multi-nozzle cooperative control system.

5. The brake drum residual stress detection and shot blasting control parameter adaptive control method for extreme temperature change environment, according to claim 1, characterized in that, The closed-loop control cycle specifically includes: Comparing the result data detected online with the expected target value to generate a deviation signal; The deviation signal is fed back to the multi-objective optimization decision model; Using an optimization algorithm to update the parameters of the multi-objective optimization decision model online to realize autonomous learning and continuous optimization of shot blasting process parameters.

6. A system for implementing the method of adaptive control of control parameters for residual stress detection and shot blasting of brake drums in extreme temperature change environments according to any one of claims 1 to 5, characterized in that, It includes: An embedded multi-source sensor detection module for real-time acquisition of state data of the brake drum under extreme temperature change environment; A stress field dynamic reconstruction and calculation module for running the thermal-mechanical-phase change three-field coupling finite element inversion algorithm; An intelligent decision and optimization module embedded with the multi-objective optimization decision model trained by machine learning; A multivariable adaptive execution control module for receiving instructions and driving the shot blasting equipment; An online quality feedback and closed-loop control module for quality verification and parameter optimization.

7. The system of claim 6, wherein, The high-temperature-resistant fiber grating sensor in the embedded multi-source sensing detection module is pre-installed at a specific depth of the brake drum by laser cladding technology and can reliably work in an environment of 600 DEG C or above.

8. The system of claim 6, wherein, The multivariable adaptive execution control module comprises: A high-speed electromagnetic valve controlled projectile flow system is used for accurately controlling the projectile injection amount per unit time; A six-degree-of-freedom robot arm is used for adjusting the injection pose and angle according to program instructions. An active temperature compensation unit is used for dynamically correcting the relationship between the shot blasting energy and the material surface response according to the real-time temperature signal.

9. The system of claim 6, wherein, The online quality feedback and closed-loop control module is integrated with a laser speckle interferometer and a microhardness tester, and can quickly detect the residual stress distribution, microhardness and surface morphology of the brake drum after shot blasting without interrupting the production rhythm.

Images