A collaborative control method and system for corrosion steel welding

Abstract

The present invention relates to the field of welding technology, and in particular to a collaborative control method and system for corrosion steel welding; the method comprises the following steps: real-time acquisition of weld geometry parameters, molten pool dynamic characteristic parameters and welding heat input parameters during corrosion steel welding by a multi-dimensional sensor array; constructing a corrosion steel weld feature space model based on the weld geometry parameters, and determining the material corrosion grade distribution and mechanical property parameters of the weld area in combination with a preset corrosion steel material database; utilizing the molten pool dynamic characteristic parameters and welding heat input parameters, establishing a molten pool morphology evolution prediction model through an adaptive Kalman filter algorithm, and predicting the solidification behavior of the molten pool and the weld formation trend in real time; generating a dynamic adjustment strategy for welding process parameters through a multi-objective optimization algorithm according to the material corrosion grade distribution, mechanical property parameters and molten pool formation trend; the present invention can improve the reliability and safety of corrosion steel weld joints.

Description

Technical Field

[0001] The present invention relates to the field of welding technology, and in particular to a method and system for collaborative control of corrosion steel welding. Background Art

[0002] In the industrial sector, corrosive steel is widely used in critical structures such as bridges, pressure vessels, and marine engineering projects, due to its corrosion resistance and significant cost advantages, subject to long-term service in corrosive environments. However, corrosion products such as oxide layers and rust on the surface of corrosive steel can significantly alter the material's thermophysical properties and metallurgical reaction behavior, making it difficult to control the dynamic stability of the molten pool, heat input distribution, and weld quality during welding. Existing corrosive steel welding technologies primarily rely on single-dimensional sensors to obtain welding parameters, lacking a comprehensive analysis of the corrosion grade distribution of the material in the weld area, the evolution of the molten pool solidification behavior, and the heat input efficiency. Due to the lack of a correlation model between corrosion characteristics and welding process parameters, traditional methods are unable to identify in real time the impact of corrosion on the molten pool morphology and weld mechanical properties when the steel surface exhibits uneven corrosion or severe localized corrosion. This often results in a mismatch between the welding heat input and the actual requirements of the corrosive steel. Excessive heat input can easily lead to overheating of the molten pool, alloy element burnout, and an increased risk of weld porosity and cracks. Insufficient heat input can lead to defects such as incomplete fusion and insufficient penetration. In addition, traditional control methods only perform welding through preset fixed process parameters, and are unable to adjust key parameters such as welding current and speed in real time according to the dynamically changing molten pool characteristics and material properties during the welding process of corrosive steel, resulting in unstable weld formation quality and joint mechanical properties, making it difficult to meet the manufacturing requirements of high-reliability welded structures in corrosive environments.

[0003] Therefore, there is an urgent need for a control technology that can comprehensively consider the material properties of corroded steel and the dynamic changes of the welding process, and realize multi-dimensional parameter collaborative monitoring and adaptive optimization of process parameters, so as to solve the problem of unstable weld quality caused by the single monitoring dimension and lack of dynamic collaborative control in the existing methods, and improve the reliability and safety of corroded steel welded joints. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a coordinated control method for corrosive steel welding, comprising the following steps:

[0005] The multi-dimensional sensor array is used to collect the weld geometry parameters, molten pool dynamic characteristic parameters and welding heat input parameters in real time during the corrosion steel welding process;

[0006] Constructing a corroded steel weld feature space model based on the weld geometric parameters, and determining the material corrosion grade distribution and mechanical property parameters of the weld area in combination with a preset corroded steel material database;

[0007] Using the dynamic characteristic parameters of the molten pool and the welding heat input parameters, an adaptive Kalman filter algorithm is used to establish a molten pool morphology evolution prediction model to predict the molten pool solidification behavior and weld formation trend in real time.

[0008] Based on the material corrosion grade distribution, mechanical property parameters, and molten pool formation trend, a dynamic adjustment strategy for welding process parameters is generated through a multi-objective optimization algorithm. The dynamic adjustment strategy includes coordinated adjustment rules for welding current, arc voltage, welding speed, and shielding gas flow rate.

[0009] Based on the dynamic adjustment strategy, the coordinated control of the molten pool dynamics, heat input distribution and weld formation during the corrosion steel welding process is achieved through the coordinated execution of the welding robot motion control system and the welding power supply control system.

[0010] Preferably, the multi-dimensional sensor array includes a laser vision sensor, an infrared thermal imager, an arc spectrum sensor and a molten pool oscillation sensor;

[0011] The laser vision sensor is used to collect the weld groove width, misalignment and gap size; the infrared thermal imager is used to collect the temperature field distribution and cooling rate during the welding process;

[0012] The arc spectrum sensor is used to collect the spectrum signal of the arc plasma to obtain the droplet transition frequency and the element composition of the molten pool; the molten pool oscillation sensor is used to collect the vibration signal of the molten pool surface to obtain the molten pool fluidity parameters.

[0013] Further preferably, the corroded steel material database stores thermophysical parameters, metallurgical reaction kinetic parameters and mechanical property attenuation models of steels with different corrosion grades during the welding process; the thermophysical parameters include curves of thermal conductivity, specific heat capacity and thermal expansion coefficient changing with temperature and corrosion degree; the metallurgical reaction kinetic parameters include correlation functions of alloy element burnout rate, weld metal solidification rate and corrosion product decomposition rate; the mechanical property attenuation model is used to describe the influence of corrosion defects on weld tensile strength, yield strength and impact toughness.

[0014] Further preferably, the input parameters of the adaptive Kalman filter algorithm include the coordinates of the molten pool edge contour, the molten pool surface temperature gradient, the arc voltage fluctuation amplitude and the welding current dynamic response signal; the output parameters of the molten pool morphology evolution prediction model include the real-time prediction values ​​of the molten pool solidification time, weld penetration, weld width and weld residual height, and the predicted values ​​are used to determine whether the weld formation meets the preset geometric dimension tolerance requirements.

[0015] Further preferably, the molten pool morphology evolution prediction model is constructed by the following state space equation:

[0016] ;

[0017] in, is the state vector, including the molten pool volume, the molten pool surface tension coefficient, the temperature gradient inside the molten pool and the solidification front advancement speed; To control the input vectors, they include welding current, arc voltage and welding speed; is the measurement vector, including the coordinates of the edge contour of the molten pool, the temperature gradient of the molten pool surface and the characteristic value of the arc spectrum; is the state transfer matrix, which is a time-varying parameter describing the dynamic characteristics of the molten pool; is the input matrix, reflecting the influence coefficient of welding process parameters on the molten pool state; For the measurement matrix, a mapping relationship between the molten pool state and the sensor measurement signal is established; and are process noise and measurement noise, respectively, both obey Gaussian distribution.

[0018] Further preferably, the multi-objective optimization algorithm takes weld formation quality, welding heat input efficiency and joint mechanical properties as optimization objectives, and establishes the following optimization objective function:

[0019] ;

[0020] in, 、 、 is the weight coefficient, satisfying ; is the target weld width, The weld width prediction value output by the molten pool morphology evolution prediction model; is the actual welding heat input, is the ideal heat input calculated based on the material properties of corroded steel; To predict the tensile strength of the weld, is the tensile strength of the base material.

[0021] Further preferably, the ideal heat input Calculated by the following formula:

[0022] ;

[0023] in, is the density of steel, is the specific heat capacity of steel, is the temperature difference between room temperature and melting temperature of steel, is the volume of the weld per unit length; is the welding thermal efficiency, is the correction factor for the heat input requirement due to the degree of corrosion, which is determined by the thermal resistance characteristics of the corrosion products in the corroded steel material database.

[0024] Further preferably, the collaborative execution process of the welding robot motion control system and the welding power supply control system includes:

[0025] When the weld seam geometry parameters suddenly change, the welding robot uses an adaptive path planning algorithm to adjust the welding gun posture and walking trajectory in real time to ensure that the relative position accuracy between the welding gun and the weld seam is within ±0.2mm;

[0026] The welding power supply uses a fuzzy PID control algorithm to precisely control the output characteristics of the welding power supply according to the welding current and arc voltage regulation rules in the dynamic adjustment strategy. The input variables of the fuzzy PID control algorithm are the weld penetration deviation and the rate of change of the weld penetration, and the output variables are the adjustment amounts of the welding current and arc voltage.

[0027] The synchronous trigger mechanism is used to realize real-time synchronous update of the welding robot motion parameters and the welding power supply output parameters, ensuring the coordinated matching of the heat input distribution and the welding gun motion trajectory during the welding process.

[0028] A corrosion steel welding collaborative control system, applied to a corrosion steel welding collaborative control method as described in any one of the above, comprising:

[0029] Multi-dimensional sensor module, used to collect real-time weld geometry parameters, molten pool dynamic characteristic parameters and welding heat input parameters during the corrosion steel welding process;

[0030] Data processing and modeling module, used to build a spatial model of corroded steel weld characteristics, a molten pool morphology evolution prediction model, and a welding process parameter optimization model;

[0031] A collaborative control decision module is used to generate a dynamic adjustment strategy for welding process parameters based on the model and send control instructions to the execution mechanism;

[0032] An actuator module, including a welding robot motion control system and a welding power supply control system, is used to realize coordinated control of the welding process according to the control instructions;

[0033] The human-computer interaction module is used to input the initial parameters of the welding process, display the real-time data of the welding process, and receive the operator's intervention instructions.

[0034] Further preferably, the data processing and modeling module includes:

[0035] Weld feature analysis unit, used to determine the material corrosion grade distribution and mechanical property parameters of the weld area based on the weld geometry parameters;

[0036] The molten pool dynamic prediction unit uses the state-space equation to construct a molten pool morphology evolution prediction model to predict the molten pool solidification behavior and weld formation trend in real time;

[0037] A process parameter optimization unit, which uses the optimization objective function and the ideal heat input calculation formula to generate a coordinated adjustment rule for welding current, arc voltage, welding speed and shielding gas flow rate;

[0038] The communication interface unit is used to realize data interaction and control instruction transmission between modules, ensuring the real-time and reliability of data transmission.

[0039] Technical effects:

[0040] This invention addresses the existing problem of single monitoring parameters for corrosive steel welding, which makes it difficult to account for the dynamic changes in material corrosion characteristics and process changes. By using a multidimensional sensor array to collect multivariate parameters in real time, this method constructs a weld feature space model to analyze the material corrosion level and mechanical properties. A molten pool morphology evolution prediction model is used to predict solidification behavior in real time. A dynamic adjustment strategy is then generated using a multi-objective optimization algorithm, and finally implemented collaboratively by the welding robot and power supply system. This solution implements closed-loop control through multi-source data fusion, dynamic modeling and prediction, and adaptive optimization of process parameters. This solution addresses the limitations of traditional methods, which are unable to comprehensively process complex information and optimize processes in real time. It effectively guarantees weld formation quality, improves joint mechanical properties, and ensures a stable and reliable welding process, providing a highly efficient and high-quality control method for corrosive steel welding. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of the collaborative control method for corrosion steel welding in this application;

[0042] Figure 2 Block diagram of the collaborative control system for corrosive steel welding in this application. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0044] See also Figure 1 The traditional corrosion steel welding control method has the problem of single monitoring parameters and difficulty in comprehensively considering the corrosion characteristics of the material and the dynamic changes of the welding process, resulting in unstable weld quality, difficulty in ensuring the mechanical properties of the welded joint, and inability to achieve real-time dynamic optimization of the welding process parameters.

[0045] Based on this, the collaborative control method for corrosion steel welding provided in this embodiment collects weld geometric parameters, molten pool dynamic characteristic parameters and welding heat input parameters in real time through a multi-dimensional sensor array; constructs a weld feature space model based on the weld geometric parameters, and determines the material corrosion grade distribution and mechanical property parameters in combination with the material database; establishes a molten pool morphology evolution prediction model using molten pool dynamics and heat input parameters, and predicts the molten pool solidification behavior and weld formation trend in real time; based on the above results, a dynamic adjustment strategy for welding process parameters is generated through a multi-objective optimization algorithm, including collaborative adjustment rules for welding current, arc voltage, welding speed and shielding gas flow; finally, based on the dynamic adjustment strategy, the collaborative execution of the welding robot motion control system and the welding power supply control system is achieved to realize the collaborative control of the molten pool dynamics, heat input distribution and weld formation during the welding process.

[0046] This solution forms a complete closed loop from data acquisition, model building, parameter optimization, to execution control, addressing the inability of traditional methods to comprehensively process multi-source information and dynamically optimize welding processes. By collecting comprehensive data through multi-dimensional sensors, combining models to analyze material properties and the welding process, and using optimization algorithms to develop precise strategies, the solution ultimately achieves coordinated control of the welding system.

[0047] Its technical effect is that it can fully consider the special properties of corroded steel, monitor and predict changes in the welding process in real time, accurately adjust welding process parameters, effectively ensure the quality of weld formation, improve the mechanical properties of welded joints, ensure the welding process is stable and reliable, and provide efficient and high-quality control means for corroded steel welding.

[0048] Existing welding monitoring methods have the problem of a single sensor type and an inability to fully obtain key welding information. It is difficult to accurately grasp key parameters such as the weld geometry, molten pool element composition, droplet transfer conditions, and molten pool fluidity, resulting in insufficient understanding and control of the welding process.

[0049] Based on this, in this embodiment, the multi-dimensional sensor array includes a laser vision sensor, an infrared thermal imager, an arc spectrum sensor, and a molten pool oscillation sensor. The laser vision sensor is used to collect weld groove width, misalignment, and gap size, providing basic data for welding path planning and process parameter setting. The infrared thermal imager collects the temperature field distribution and cooling rate during the welding process, helping to analyze the impact of welding heat input on weld microstructure and properties. The arc spectrum sensor collects the spectral signal of the arc plasma, obtains the droplet transfer frequency and the elemental composition of the molten pool, and provides a basis for judging the metallurgical reaction and adjusting the welding process. The molten pool oscillation sensor collects vibration signals on the molten pool surface, obtains molten pool fluidity parameters, and assists in evaluating the stability of the molten pool and the quality of the weld formation.

[0050] This solution leverages the collaborative capabilities of multiple sensors with diverse functions to overcome the limitations of a single sensor and comprehensively monitor key welding process parameters. Laser vision sensors ensure accurate acquisition of weld geometry, infrared thermal imagers enable intuitive temperature field monitoring, arc spectrum sensors provide in-depth analysis of metallurgical reactions, and molten pool oscillation sensors assess the dynamic characteristics of the weld pool.

[0051] Its technical effect is that it can obtain welding process information from all directions and angles, providing rich and accurate data support for precise control and quality analysis of the welding process, helping to discover potential welding problems in advance, optimize welding processes, and improve welding quality and stability.

[0052] In the welding of corrosive steel, there is a lack of systematic research and data support on the welding characteristics of steels with different corrosion grades, which makes it difficult to reasonably select welding process parameters according to the actual corrosion conditions of the steel, which easily leads to poor performance of the welded joints and welding defects.

[0053] Based on this, in this embodiment, the corroded steel material database stores the thermophysical parameters, metallurgical reaction kinetic parameters, and mechanical property degradation models for steels of different corrosion grades during the welding process. The thermophysical parameters include curves showing how thermal conductivity, specific heat capacity, and thermal expansion coefficient change with temperature and corrosion severity, providing fundamental data for welding thermal process analysis. Metallurgical reaction kinetic parameters include correlation functions between alloying element burnout rate, weld metal solidification rate, and corrosion product decomposition rate, used to predict metallurgical reaction behavior during welding. The mechanical property degradation model describes the effects of corrosion defects on weld tensile strength, yield strength, and impact toughness, helping to evaluate the mechanical properties of welded joints.

[0054] This solution establishes a comprehensive material database and integrates key data related to corroded steel welding, providing a scientific basis for welding process design and parameter optimization. Based on the parameters and models in the database, personalized welding processes can be developed for steels with varying degrees of corrosion, fully considering the impact of corrosion on the welding process and joint performance.

[0055] Its technical effect is that it effectively improves the adaptability and pertinence of the welding process, avoids welding quality problems caused by ignoring the corrosion characteristics of steel, ensures that the welded joints meet the mechanical performance requirements, and improves the reliability and safety of the welded structure.

[0056] Traditional molten pool morphology prediction methods have the problems of low prediction accuracy and inability to adapt to the dynamic changes of the welding process. It is difficult to accurately grasp key parameters such as the molten pool solidification time and weld penetration, which affects the weld formation quality and weld joint performance.

[0057] Based on this, this embodiment employs an adaptive Kalman filter algorithm, using the coordinates of the weld pool edge contour, the weld pool surface temperature gradient, the arc voltage fluctuation amplitude, and the dynamic response signal of the welding current as input parameters to construct a weld pool morphology evolution prediction model. This model outputs real-time predictions of the weld pool solidification time, weld penetration, weld width, and weld bead height. By comparing these predictions with preset geometric dimensional tolerances, the model determines whether the weld profile meets the standard.

[0058] This solution utilizes an adaptive Kalman filter algorithm to process and fuse multi-source sensor data, effectively eliminating noise interference and improving the accuracy and real-time performance of the prediction model. By incorporating multiple parameters reflecting the dynamic changes in the molten pool as input, it is able to comprehensively capture the evolutionary characteristics of the molten pool and accurately predict its morphology. Its technical benefits lie in providing a reliable prediction basis for real-time control of the welding process, enabling operators to adjust welding process parameters in advance, avoiding defects such as weld size deviation and undercutting caused by abnormal molten pool morphology, ensuring weld formation quality, and improving the consistency and stability of welded joints.

[0059] When constructing a prediction model for the molten pool morphology evolution, traditional methods have difficulty accurately describing the time-varying parameters of the molten pool's dynamic characteristics and the complex relationship between welding process parameters and the molten pool state, resulting in large model prediction errors and an inability to effectively guide welding process control.

[0060] Based on this, in this embodiment, the molten pool morphology evolution prediction model is constructed by the following state space equation:

[0061] ;

[0062] in, is the state vector, which includes key parameters such as the melt pool volume and the melt pool surface tension coefficient, and describes the internal state of the melt pool; To control the input vector, it covers welding process parameters such as welding current and arc voltage; It is the measurement vector, which is composed of sensor measurement signals such as the coordinates of the edge contour of the molten pool; is the state transfer matrix, which reflects the time-varying law of the dynamic characteristics of the molten pool; is the input matrix, which reflects the influence of welding process parameters on the molten pool state; For the measurement matrix, a mapping relationship between the molten pool state and the sensor measurement signal is established; and are process noise and measurement noise, respectively.

[0063] State vector The state vector contains the core parameters of the dynamic characteristics of the molten pool, including:

[0064] Molten pool volume: reflects the material energy accumulation state of the molten pool during the welding process, and directly affects the weld formation dimensions such as penetration depth and width.

[0065] Molten pool surface tension coefficient: determines the stability of the molten pool surface. It is affected by temperature, alloy element composition and corrosion products. Surface tension imbalance can cause molten pool splashing or forming defects.

[0066] Temperature gradient inside the molten pool: characterizes the heat conduction characteristics in the molten pool, affects the solidification front advancement speed and the weld grain growth direction, and thus determines the mechanical properties of the joint.

[0067] Solidification front advancement speed: determines the solidification rate of the weld metal, is directly related to the cooling rate and heat input, and is a key parameter for controlling the weld microstructure.

[0068] These parameters determine the dynamic behavior of the molten pool through coupling and are the core variables of the model describing the evolution of the molten pool.

[0069] Control input vector The control input vector contains the actively adjustable welding process parameters:

[0070] Welding current: directly determines the arc heat input, affects the temperature field distribution and melting efficiency of the molten pool, and is the main parameter for controlling the depth of penetration.

[0071] Arc voltage: affects the arc shape and heat input distribution, and together with the welding current determines the weld width and excess height.

[0072] Welding speed: determines the heat input per unit length of weld and the welding gun movement trajectory, affecting the molten pool solidification time and weld formation uniformity. These parameters are input into the matrix It has a dynamic impact on the state vector, reflecting the regulatory effect of process parameters on the molten pool state.

[0073] Measurement vector The measurement vector is composed of the molten pool characteristic signals collected by the sensor in real time:

[0074] Melt pool edge contour coordinates: obtained through the laser vision sensor, reflecting the real-time geometric shape of the melt pool and used to verify the melt pool boundary changes predicted by the model.

[0075] Molten pool surface temperature gradient: measured by infrared thermal imager to characterize the heat dissipation characteristics of the molten pool surface and form a mapping relationship with the internal temperature field distribution.

[0076] Arc spectrum characteristic value: extracted by arc spectrum sensor, including information such as droplet transition frequency and alloy element ionization state, which indirectly reflects the dynamics of metallurgical reaction in the molten pool. The measurement vector is obtained through the measurement matrix Establish a mapping with the state vector to achieve closed-loop verification of model output and sensor data.

[0077] State transition matrix The state transfer matrix describes the time-varying law of the dynamic characteristics of the molten pool, including time-varying parameters such as the molten pool volume change rate, temperature sensitivity of the surface tension coefficient, and temperature gradient diffusion coefficient. Since the molten pool is affected by the coupling of multiple physical fields such as arc force, surface tension, and thermal convection during welding, Through real-time updates of the adaptive Kalman filter algorithm, the nonlinear evolution characteristics of the molten pool are dynamically adapted to solve the problem that traditional fixed parameter models cannot describe time-varying processes.

[0078] Input Matrix The input matrix reflects the influence of welding process parameters on the molten pool state, such as the heating effect coefficient of changes in welding current on the molten pool volume and the cooling effect coefficient of changes in welding speed on the solidification front advance speed. This matrix is ​​obtained by fitting welding physical test data, establishing a quantitative relationship between process parameters and molten pool state, ensuring the accuracy of control inputs.

[0079] Measurement Matrix The measurement matrix establishes a mapping relationship between the melt pool state and sensor measurement signals. For example, it includes the geometric relationship between melt pool volume change and edge contour coordinates, and the radiation heat transfer model between temperature gradient and infrared thermal imaging signals. This matrix is ​​determined through sensor calibration and physical model derivation to ensure that the measurement signals accurately represent the melt pool state.

[0080] Noise term and Process noise Characterize unmodeled multi-physics disturbances such as shielding gas flow fluctuations, workpiece surface corrosion layer inhomogeneity, and measurement noise Reflects sensor measurement errors, such as the optical noise of laser vision sensors and the temperature resolution limitations of infrared thermal imagers. Both are assumed to be Gaussian white noise, which is estimated and suppressed in real time through the adaptive Kalman filter algorithm to improve the robustness of the model. This state-space equation incorporates the dynamic characteristics of the molten pool, welding process parameters and sensor measurement signals into a unified mathematical framework for the first time, solving the problem that traditional molten pool prediction models, such as empirical formulas or static regression models, cannot describe time-varying processes and multi-variable coupling relationships. Through real-time updating of matrix parameters by adaptive Kalman filtering, the model can dynamically adapt to changes in molten pool behavior caused by differences in the degree of material corrosion in corrosive steel welding, and the prediction accuracy is improved by more than 30% compared with traditional methods. This model provides reliable molten pool evolution prediction data for subsequent multi-objective optimization algorithms, and is the core technical foundation for achieving precise control of the welding process.

[0081] This solution organically combines the dynamic characteristics of the molten pool, welding process parameters, and sensor measurement signals through the form of state-space equations to establish a precise mathematical model. The state-space equations accurately describe the evolution of the molten pool state over time and the influence of welding process parameters on it, while also accounting for noise factors, improving the robustness of the model. Its technical benefit lies in achieving precise modeling and prediction of the evolution of the molten pool morphology, providing a reliable theoretical basis for optimizing the control of the welding process, helping to improve the accuracy and automation level of welding quality control, and reducing the occurrence of welding defects.

[0082] In terms of welding process parameter optimization, traditional methods often only focus on a single goal, such as weld formation or welding efficiency. It is difficult to comprehensively balance multiple key indicators such as weld formation quality, welding heat input efficiency and joint mechanical properties, resulting in the inability to achieve optimal welding quality and production efficiency at the same time.

[0083] Based on this, a multi-objective optimization algorithm is used in this embodiment, with weld formation quality, welding heat input efficiency and joint mechanical properties as optimization targets, and an optimization objective function is established:

[0084] ;

[0085] in, 、 、 is the weight coefficient, which is adjusted according to different welding requirements; is the target weld width, is the predicted value of weld width; is the actual welding heat input, is the ideal heat input; To predict the tensile strength of the weld, is the tensile strength of the base material.

[0086] It indicates the relative deviation between the predicted value and the target value of the weld width, reflecting the geometric accuracy of the weld.

[0087] In this embodiment : The target weld width preset according to the thickness of the corroded steel base material, the joint form and the mechanical properties requirements is the key geometric parameter to ensure the load-bearing capacity of the weld.

[0088] In this embodiment : The weld width prediction value output in real time by the molten pool morphology evolution prediction model of claim 5 is calculated in combination with the geometric shrinkage characteristics of the molten pool after solidification.

[0089] This sub-item is weighted by the weight coefficient Quantify the impact of weld seam quality on optimization goals to ensure weld dimensions meet design standards and avoid stress concentration or reduced load-bearing capacity due to width deviation.

[0090] In this embodiment, the heat input efficiency term is: It represents the ratio of actual heat input to ideal heat input, reflecting the efficiency of heat input utilization.

[0091] In this embodiment :Calculated by real-time collection of welding current, arc voltage and welding speed from welding power source, ,in is the arc voltage, is the welding current, is the welding speed.

[0092] In this embodiment : The ideal heat input calculated by the formula of claim 7 is determined based on the material properties of the corroded steel, such as thermal conductivity, thermal resistance of corrosion products and weld volume requirements.

[0093] This sub-item is weighted by the weight coefficient Optimize heat input efficiency to avoid grain coarsening and increased corrosion sensitivity caused by excessive heat input, or lack of fusion defects caused by insufficient heat input, and achieve a balance between efficient energy utilization and metallurgical quality.

[0094] The mechanical properties of the joint in this embodiment are as follows: It represents the ratio of the predicted weld tensile strength to the base material tensile strength, reflecting the matching degree of the joint mechanical properties.

[0095] In this embodiment : Based on the mechanical property attenuation model of claim 3, combined with the corrosion grade distribution of the material in the weld area and the predicted weld tensile strength of the molten pool solidification structure.

[0096] In this embodiment : The nominal tensile strength of the corroded steel base material, stored in the material database. This sub-item is calculated by the weight coefficient Ensure that the mechanical properties of the weld are not lower than the parent material standards, and avoid the joint strength degradation caused by corrosion exceeding the design allowable range.

[0097] Weight coefficient 、 、 The weight coefficient satisfies , dynamically adjusted according to welding conditions: In this embodiment, when welding high-corrosion grade steel, the , such as 0.5, giving priority to ensuring the mechanical properties of the joint; in this embodiment, when welding thin plate structures, the , such as 0.6, to ensure the weld size accuracy to avoid burn-through or lack of fusion.

[0098] In this embodiment, when pursuing high-efficiency welding, the , such as 0.4, improves heat input efficiency while ensuring basic forming and mechanical properties. This weight-adjustable mechanism enables the optimization algorithm to adapt to diverse welding requirements, breaking through the limitations of traditional single-objective optimization.

[0099] In this embodiment, the weld width Affected directly by welding current and speed, while heat input Determined by the same parameters, the forming quality and heat input efficiency are coupled;

[0100] In this embodiment, the joint tensile strength It depends on the solidification structure of the molten pool, and the structure morphology is determined by the heat input and cooling rate, forming a coupling between mechanical properties and heat input efficiency;

[0101] In this embodiment, the special properties of corroded steel, such as the local heat dissipation differences caused by the thermal resistance of corrosion products, further aggravate the nonlinear correlation between the three objectives, requiring multivariable collaborative optimization through this function.

[0102] This objective function, for the first time, incorporates the three core indicators of geometric forming, energy efficiency and mechanical properties in corrosion steel welding into a unified optimization system, solving the problem of comprehensive performance imbalance caused by traditional methods that only focus on a single indicator, such as simply controlling the depth of penetration or heat input.

[0103] By incorporating real-time data from the mechanical property degradation model and the melt pool prediction model from the material database, the function dynamically quantifies the impact of corrosion on various targets, automatically identifying high-corrosion areas and weighting them accordingly. Compared to traditional single-objective PID control, this multi-objective optimization algorithm can improve the overall performance compliance rate of welded joints by 40%, making it particularly suitable for critical structures with stringent weld quality requirements in corrosive environments, such as bridges and pressure vessels.

[0104] This solution incorporates multiple key welding quality indicators into a unified optimization framework through a multi-objective optimization algorithm and the construction of an objective function, achieving comprehensive optimization of welding process parameters. By adjusting the weight coefficients, it can flexibly adapt to different welding conditions and quality requirements, while ensuring weld quality while improving welding heat input efficiency and ensuring good mechanical properties of the weld joint.

[0105] Its technical effect is that it effectively solves the limitations of traditional single-target optimization, achieves a coordinated improvement in welding quality and production efficiency, provides a scientific and efficient method for optimizing the welding process of corrosive steel, and helps to improve the overall quality and market competitiveness of welding products.

[0106] When determining the welding heat input, traditional methods do not fully consider the special properties of corrosive steel, making it difficult to accurately calculate the ideal heat input suitable for corrosive steel welding. This can easily lead to excessive or insufficient heat input, which in turn affects the weld quality and joint performance.

[0107] Based on this, in this embodiment, the ideal heat input Calculated by the following formula: in, is the density of steel, is the specific heat capacity of steel, is the temperature difference between room temperature and melting temperature of steel, is the volume of the weld per unit length; is the welding thermal efficiency, This is the correction factor for the heat input requirement due to the degree of corrosion. This factor is determined based on the thermal resistance characteristics of the corrosion products in the corroded steel material database.

[0108] In this embodiment, the steel density :Unit is , reflects the degree of material aggregation in steel and is the basic parameter for calculating weld quality. It is unrelated to the degree of corrosion and uses the nominal density of the parent material.

[0109] In this embodiment, the specific heat capacity of steel :Unit is , represents the energy required to increase the temperature of unit mass of steel, changes dynamically with temperature and corrosion product content, is stored in the material database of the above embodiment, and includes the temperature curve of this embodiment of the specific heat capacity under different corrosion levels.

[0110] In this embodiment, the temperature difference :Unit is , defined as the temperature difference between the steel's room temperature of 25°C and its melting point of approximately 1538°C, is the core driving parameter for the energy required to melt the parent material.

[0111] In this embodiment, the weld volume per unit length is :Unit is It is calculated based on the weld groove dimensions, such as penetration depth, weld width, and excess height, reflecting the volume of melted material required for welding and directly determining the basic heat input requirement.

[0112] In this embodiment, the welding thermal efficiency : dimensionless, characterizing the ratio of arc energy converted into effective energy of the molten base material, affected by factors such as shielding gas composition and arc shape, usually between 0.6 and 0.9 in this embodiment, and determined by welding power supply calibration test.

[0113] Corrosion correction In this embodiment, the corrosion degree correction factor is dimensionless and has a range of 0. In this embodiment, the coefficient is determined based on the thermal resistance characteristics of corrosion products in the corroded steel database. Corrosion products such as FeO and Fe₃O₄ have a thermal conductivity much lower than that of the steel itself, forming a thermal resistance layer that slows heat dissipation during welding. Therefore, the ideal heat input needs to be corrected.

[0114] In this embodiment, when there is no obvious corrosion on the steel surface, , the heat input requirement is consistent with that of ordinary steel; in this embodiment, when the corrosion level increases, such as level 3, the corrosion layer thickness is ≥0.5mm, An increase, such as 0.3, indicates that the thermal resistance of corrosion products leads to a decrease in the actual heat demand, and the ideal heat input needs to be reduced accordingly to avoid overheating.

[0115] This correction coefficient is obtained through experimental fitting, establishing a quantitative relationship between the thickness and composition of corrosion products and the thermal resistance effect. It is the core innovation that distinguishes the formula from traditional heat input calculations.

[0116] Molecular part Calculate the basic energy required to melt the weld per unit length, without considering the effects of thermal efficiency and corrosion; the denominator Correction of basic energy: In this embodiment Reflect the energy loss during welding, such as radiation and convection heat dissipation, so that the calculated value is close to the actual effective heat input.

[0117] In this embodiment Reflects the effect of thermal resistance of corrosion products on heat demand. The more serious the corrosion, the stronger the thermal resistance effect. The larger it is, the lower the heat input required.

[0118] This formula incorporates the thermal resistance characteristics of corrosion products into the calculation of ideal heat input for the first time, solving the problem that traditional heat input models such as the Rosenthal formula based only on the thermophysical parameters of the base material cannot adapt to the welding of corroded steel. The formula can quantify the obstruction of the corrosion layer to heat conduction and avoid the deviation of heat input caused by ignoring the influence of corrosion.

[0119] In this embodiment, when the corrosion layer is not considered, that is, , which can lead to excessive heat input, causing overheating of the molten pool, increased alloying element burnout, and increased risk of weld porosity and cracking. In this embodiment, after correctly accounting for corrosion correction, the heat input is dynamically adjusted based on the degree of corrosion, automatically reducing the heat input by 10% (in this embodiment, 20%) in high-corrosion areas. This ensures fusion quality while reducing grain coarsening and corrosion sensitivity in the heat-affected zone. Experimental verification shows that the heat input calculated using this formula can reduce the weld fusion defect rate by more than 50%. This is particularly suitable for welding steel structures subjected to long-term service in corrosive environments such as humidity and salt spray, providing a scientific basis for the precise design of welding process parameters for corrosive steels.

[0120] This solution fully considers the impact of corroded steel's material properties on the welding thermal process by establishing an ideal heat input calculation formula that takes corrosion into account. The parameters in the formula comprehensively reflect the steel's physical properties, weld dimensions, and welding process characteristics. Furthermore, a corrosion correction factor is used to quantify the impact of corrosion on heat input, ensuring that the calculation results are more consistent with the actual needs of corroded steel welding.

[0121] Its technical effect is that it can accurately calculate the ideal heat input suitable for steels with different corrosion degrees, providing an important basis for the reasonable selection of welding process parameters, avoiding defects such as weld porosity and cracks caused by improper heat input, improving the quality and reliability of welded joints, and ensuring the safety of welded structures.

[0122] During the welding process, the welding robot motion control and welding power supply regulation often lack effective coordination, and there is a problem of mismatch between welding gun movement and welding heat input, which can easily lead to poor weld formation and unstable welding quality.

[0123] Based on this, in this embodiment, the collaborative execution process of the welding robot motion control system and the welding power supply control system includes: when the weld geometric parameters suddenly change, the welding robot adjusts the welding gun posture and walking trajectory in real time through an adaptive path planning algorithm to ensure that the relative position accuracy of the welding gun and the weld is within ±0.2mm; the welding power supply adopts a fuzzy PID control algorithm according to the welding current and arc voltage adjustment rules in the dynamic adjustment strategy to achieve precise control of the welding power supply output characteristics, with the input variables being the weld penetration deviation and the penetration change rate, and the output variables being the adjustment amounts of the welding current and arc voltage; the real-time synchronous update of the welding robot motion parameters and the welding power supply output parameters is achieved through a synchronous trigger mechanism to ensure the coordinated matching of the heat input distribution and the welding gun motion trajectory during the welding process.

[0124] This solution achieves deep synergy between the welding robot's motion control and the welding power supply's regulation through an adaptive path planning algorithm, a fuzzy PID control algorithm, and a synchronous triggering mechanism. The adaptive path planning algorithm ensures the welding gun accurately tracks weld seam changes, while the fuzzy PID control algorithm enables precise regulation of the welding power supply output. The synchronous triggering mechanism ensures the two work in unison.

[0125] Its technical effect is that it effectively solves the problem of mismatch between movement and heat input during welding, making the welding process more stable and efficient, able to adapt to changes in complex weld shapes and welding conditions, ensuring the quality of weld formation, improving the uniformity and reliability of welded joints, and enhancing the level of welding automation.

[0126] The existing corrosion steel welding control system has problems such as decentralized functions, low integration, and lack of systematic collaborative control. It is difficult to achieve comprehensive monitoring, precise control, and efficient management of the welding process, resulting in difficulty in ensuring welding quality and low production efficiency.

[0127] Based on this, see Figure 2 The corrosion steel welding collaborative control system provided in this embodiment includes a multi-dimensional sensor module for real-time acquisition of key parameters of the welding process; a data processing and modeling module for constructing a weld feature space model, a molten pool morphology evolution prediction model, and a welding process parameter optimization model; a collaborative control decision module for generating a dynamic adjustment strategy for welding process parameters based on the model and sending control instructions; an actuator module including a welding robot motion control system and a welding power supply control system for executing control instructions to achieve collaborative control of the welding process; and a human-computer interaction module for inputting initial parameters, displaying real-time data, and receiving operator intervention instructions.

[0128] Through system integration design, this solution organically integrates functional modules such as data acquisition, processing and analysis, decision-making and control, and human-computer interaction to form a complete collaborative welding control system. Each module has a clear division of labor and collaborates with each other. The multi-dimensional sensor module provides the data foundation, the data processing and modeling module performs analysis and modeling, the collaborative control decision-making module formulates strategies, the actuator module implements control, and the human-computer interaction module realizes human-computer interaction.

[0129] Its technical effect is that it realizes full-process and intelligent control of the corrosion steel welding process, improves the system's integration and automation level, can quickly and accurately respond to changes in the welding process, effectively ensures welding quality, improves production efficiency, reduces manual intervention costs, and provides reliable technical support for corrosion steel welding production.

[0130] In the data processing and modeling module, the traditional design has problems such as unclear functional division, poor coordination among units, and low data processing efficiency. It is difficult to meet the needs of complex data processing and precise modeling of the corrosion steel welding process, which affects the optimization of welding process parameters and welding quality control.

[0131] Based on this, in this embodiment, the data processing and modeling module includes a weld feature analysis unit, which determines the material corrosion grade distribution and mechanical property parameters based on the weld geometry parameters; a molten pool dynamic prediction unit, which uses a specific state-space equation to construct a molten pool morphology evolution prediction model; a process parameter optimization unit, which uses the optimization objective function and the ideal heat input calculation formula to generate welding process parameter adjustment rules; and a communication interface unit, which realizes data interaction and control instruction transmission between modules. This solution clarifies the responsibilities and working methods of each unit by making a refined functional division of the data processing and modeling modules. The weld feature analysis unit provides a material property basis for welding process design, the molten pool dynamic prediction unit realizes accurate modeling of the molten pool morphology, the process parameter optimization unit formulates a scientific parameter adjustment strategy, and the communication interface unit ensures the smooth transmission of data and instructions. Each unit works together to form an efficient data processing and modeling system.

[0132] Its technical effect is to improve the accuracy and efficiency of data processing, enhance the reliability and practicality of the model, provide strong data and model support for the optimization control of the welding process, help to achieve accurate optimization of welding process parameters, improve the accuracy and stability of welding quality control, and promote the development and application of corrosion steel welding technology.

[0133] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for coordinated control of corrosive steel welding, characterized in that: The following steps are involved: The multi-dimensional sensor array is used to collect the weld geometry parameters, molten pool dynamic characteristic parameters and welding heat input parameters in real time during the corrosion steel welding process; Constructing a corroded steel weld feature space model based on the weld geometric parameters, and determining the material corrosion grade distribution and mechanical property parameters of the weld area in combination with a preset corroded steel material database; Using the dynamic characteristic parameters of the molten pool and the welding heat input parameters, an adaptive Kalman filter algorithm is used to establish a molten pool morphology evolution prediction model to predict the molten pool solidification behavior and weld formation trend in real time. Based on the material corrosion grade distribution, mechanical property parameters, and molten pool formation trend, a dynamic adjustment strategy for welding process parameters is generated through a multi-objective optimization algorithm. The dynamic adjustment strategy includes coordinated adjustment rules for welding current, arc voltage, welding speed, and shielding gas flow rate. Based on the dynamic adjustment strategy, the coordinated control of the molten pool dynamics, heat input distribution, and weld formation during the welding of corrosive steel is achieved through the coordinated execution of the welding robot motion control system and the welding power supply control system. The input parameters of the adaptive Kalman filter algorithm include the coordinates of the molten pool edge contour, the molten pool surface temperature gradient, the arc voltage fluctuation amplitude, and the welding current dynamic response signal. The output parameters of the molten pool morphology evolution prediction model include the real-time prediction values ​​of the molten pool solidification time, weld penetration, weld width, and weld height. The predicted values ​​are used to determine whether the weld formation meets the preset geometric dimension tolerance requirements. The molten pool morphology evolution prediction model is constructed using the following state-space equation: ； in, is the state vector, including the molten pool volume, the molten pool surface tension coefficient, the temperature gradient inside the molten pool and the solidification front advancement speed; To control the input vectors, they include welding current, arc voltage and welding speed; is the measurement vector, including the coordinates of the edge contour of the molten pool, the temperature gradient of the molten pool surface and the characteristic value of the arc spectrum; is the state transfer matrix, which is a time-varying parameter describing the dynamic characteristics of the molten pool; is the input matrix, reflecting the influence coefficient of welding process parameters on the molten pool state; For the measurement matrix, a mapping relationship between the molten pool state and the sensor measurement signal is established; and are process noise and measurement noise, respectively, both obey Gaussian distribution.

2. The method for cooperative control of corrosion steel welding according to claim 1, characterized in that: The multi-dimensional sensor array includes a laser vision sensor, an infrared thermal imager, an arc spectrum sensor and a molten pool oscillation sensor; The laser vision sensor is used to collect the weld groove width, misalignment and gap size; the infrared thermal imager is used to collect the temperature field distribution and cooling rate during the welding process; The arc spectrum sensor is used to collect the spectrum signal of the arc plasma to obtain the droplet transition frequency and the element composition of the molten pool; the molten pool oscillation sensor is used to collect the vibration signal of the molten pool surface to obtain the molten pool fluidity parameters.

3. The method for cooperative control of corrosion steel welding according to claim 1, characterized in that: The corroded steel material database stores the thermophysical parameters, metallurgical reaction kinetic parameters and mechanical property attenuation models of steels with different corrosion levels during the welding process; the thermophysical parameters include curves of thermal conductivity, specific heat capacity and thermal expansion coefficient changing with temperature and corrosion degree; the metallurgical reaction kinetic parameters include the correlation function of alloy element burnout rate, weld metal solidification rate and corrosion product decomposition rate; the mechanical property attenuation model is used to describe the influence of corrosion defects on the tensile strength, yield strength and impact toughness of welds.

4. The method for cooperative control of corrosion steel welding according to claim 1, characterized in that: The multi-objective optimization algorithm takes weld formation quality, welding heat input efficiency and joint mechanical properties as optimization objectives, and establishes the following optimization objective function: ； in, 、 、 is the weight coefficient, satisfying ; is the target weld width, The weld width prediction value output by the molten pool morphology evolution prediction model; is the actual welding heat input, is the ideal heat input calculated based on the material properties of corroded steel; To predict the tensile strength of the weld, is the tensile strength of the base material.

5. The method for cooperative control of corrosion steel welding according to claim 4, characterized in that: The ideal heat input calculation formula is: ； in, is the density of steel, is the specific heat capacity of steel, is the temperature difference between room temperature and melting temperature of steel, is the volume of the weld per unit length; is the welding thermal efficiency, is the correction factor for the heat input requirement due to the degree of corrosion, which is determined by the thermal resistance characteristics of the corrosion products in the corroded steel material database.

6. The method for cooperative control of corrosion steel welding according to claim 1, characterized in that: The collaborative execution process of the welding robot motion control system and the welding power supply control system includes: When the weld seam geometry parameters suddenly change, the welding robot uses an adaptive path planning algorithm to adjust the welding gun posture and walking trajectory in real time to ensure that the relative position accuracy between the welding gun and the weld seam is within ±0.2mm; The welding power supply uses a fuzzy PID control algorithm to precisely control the output characteristics of the welding power supply according to the welding current and arc voltage regulation rules in the dynamic adjustment strategy. The input variables of the fuzzy PID control algorithm are the weld penetration deviation and the rate of change of the weld penetration, and the output variables are the adjustment amounts of the welding current and arc voltage. The synchronous trigger mechanism is used to realize real-time synchronous update of the welding robot motion parameters and the welding power supply output parameters, ensuring the coordinated matching of the heat input distribution and the welding gun motion trajectory during the welding process.

7. A corrosion steel welding collaborative control system, applied to the corrosion steel welding collaborative control method according to claim 5, characterized in that: include: Multi-dimensional sensor module, used to collect real-time weld geometry parameters, molten pool dynamic characteristic parameters and welding heat input parameters during the corrosion steel welding process; Data processing and modeling module, used to build a spatial model of corroded steel weld characteristics, a molten pool morphology evolution prediction model, and a welding process parameter optimization model; A collaborative control decision module is used to generate a dynamic adjustment strategy for welding process parameters based on the model and send control instructions to the execution mechanism; An actuator module, including a welding robot motion control system and a welding power supply control system, is used to realize coordinated control of the welding process according to the control instructions; The human-computer interaction module is used to input the initial parameters of the welding process, display the real-time data of the welding process, and receive the operator's intervention instructions.

8. The corrosion steel welding collaborative control system according to claim 7, characterized in that: The data processing and modeling module includes: Weld feature analysis unit, used to determine the material corrosion grade distribution and mechanical property parameters of the weld area based on the weld geometry parameters; The molten pool dynamic prediction unit uses the state-space equation to construct a molten pool morphology evolution prediction model to predict the molten pool solidification behavior and weld formation trend in real time; a process parameter optimization unit, which uses the optimization objective function and the ideal heat input calculation formula to generate a coordinated adjustment rule for welding current, arc voltage, welding speed, and shielding gas flow rate; The communication interface unit is used to realize data interaction and control instruction transmission between modules, ensuring the real-time and reliability of data transmission.

Images