Robot automatic operation self-adaption method and system aiming at dynamic working conditions

Abstract

The invention provides a robot automatic operation self-adaption method and system for a dynamic working condition, and the method comprises the steps: anchoring a working condition evolution track, and fusing a historical working condition evolution rule with a current real-time working condition flow, so as to represent a working condition dynamic change trend; secondly, generating a work behavior gene segment, and extracting and combining work behavior core elements matched with working conditions; and then driving the operation behaviors to self-assemble, and forming a dynamic operation behavior sequence in combination with the real-time change of working conditions. In the execution process, behavior and working condition covariant signals are captured, and covariant information of operation behaviors and working conditions is captured. And finally, iteratively optimizing a baseline and a behavior scheme, reversely injecting a covariant signal into a working condition evolution trajectory generation logic, optimizing an operation behavior gene segment evolution rule, and outputting a target operation behavior scheme adaptive to the current dynamic working condition. According to the method, the operation self-adaptive capability of the robot under the dynamic working condition is remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of robots, in particular to a robot automation operation adaptive method and system for dynamic working conditions. BACKGROUND

[0002] In the field of robot automation operation, dynamic working conditions are common and complex. Traditional robot operation methods are often based on preset fixed modes and are difficult to cope with the dynamic changes of working conditions with operation progress. In the prior art, some methods can only make simple adjustments according to single real-time working condition data, and lack sufficient use of historical working condition evolution rules, resulting in inability to accurately predict future changes in working conditions. Some other methods do not fully consider the deep correlation between working conditions and operation behaviors when generating operation behaviors, so that the generated operation behaviors are difficult to accurately adapt to dynamic working conditions, and problems such as low operation efficiency and uncoordinated motion are prone to occur. Especially in complex and variable operation environments, the existing robot automation operation methods are difficult to achieve efficient, stable and adaptive operation, and cannot meet the requirements of flexibility and intelligence of robots in actual production. SUMMARY

[0003] In view of the above-mentioned problems, in combination with the first aspect of the present application, the embodiments of the present application provide a robot automation operation adaptive method for dynamic working conditions, the method comprising: Anchoring the working condition evolution trajectory, fusing the historical working condition evolution rule and the current real-time working condition flow to generate the working condition evolution trajectory, the working condition evolution trajectory representing the dynamic change trend of the working condition with the operation progress; Generating operation behavior gene fragments, extracting adaptive operation behavior core elements based on the working condition evolution trajectory, and combining to form operation behavior gene fragments, the operation behavior gene fragments carrying basic evolution information of operation actions; Driving operation behavior self-assembly, based on the operation behavior gene fragments, combining the real-time change trend of the working condition evolution trajectory, and self-assembling to form a dynamic operation behavior sequence; Capturing behavior working condition covariation signals, synchronously capturing the covariation information of operation behavior state changes and working condition evolution trajectory during the execution process of the dynamic operation behavior sequence, and forming behavior working condition covariation signals; Iterative optimization of baseline and behavior scheme, reversely injecting the behavior working condition covariation signals into the generation logic of the working condition evolution trajectory, synchronously optimizing the evolution rules of the operation behavior gene fragments, and outputting a target operation behavior scheme adapted to the current dynamic working condition.

[0004] In still another aspect, the embodiments of the present application also provide a robot automatic operation adaptive system for dynamic working conditions, comprising a processor, a machine readable storage medium, the machine readable storage medium being connected with the processor, the machine readable storage medium being used for storing programs, instructions or codes, and the processor being used for executing the programs, instructions or codes in the machine readable storage medium to realize the method described above.

[0005] Based on the above aspects, the embodiments of the present application can accurately grasp the dynamic change trend of the working condition along with the operation process by anchoring the working condition evolution track, fusing historical and real-time working condition information, and then generating an operation behavior gene segment to carry basic evolution information of operation actions, so that the operation behavior has the characteristics of being combinable and adjustable. The operation behavior is driven to self-assemble to form a dynamic operation behavior sequence in combination with the real-time change trend of the working condition, so as to ensure that the operation behavior can quickly respond to the change of the working condition. The behavior-working condition covariant signal is captured to deeply mine the cooperative change relationship between the operation behavior and the working condition. Finally, the baseline and the behavior scheme are iteratively optimized to realize high adaptation of the operation behavior scheme to the dynamic working condition. Overall, the present application significantly improves the flexibility, accuracy and adaptive ability of the robot in automatic operation under dynamic working conditions, and effectively solves the shortcomings of the traditional method in dynamic working condition operation. BRIEF DESCRIPTION OF DRAWINGS

[0006] Figure 1 is the execution flow diagram of the robot automatic operation adaptive method for dynamic working conditions provided by the embodiments of the present application.

[0007] Figure 2 is the schematic diagram of exemplary hardware and software components of the robot automatic operation adaptive system for dynamic working conditions provided by the embodiments of the present application. DETAILED DESCRIPTION

[0008] The present application will be described in detail below with reference to the accompanying drawings, Figure 1 is the flow diagram of the robot automatic operation adaptive method for dynamic working conditions provided by an embodiment of the present application, and the robot automatic operation adaptive method for dynamic working conditions will be described in detail below.

[0009] Step S110: anchoring the working condition evolution track, fusing the historical working condition evolution rule and the current real-time working condition flow to generate a working condition evolution track, and the working condition evolution track represents the dynamic change trend of the working condition along with the operation process.

[0010] In the present embodiment, the generation of the working condition evolution track is the basis for the robot to realize adaptive operation, and the core is to combine the rule of working condition change in historical welding operation and the real-time data of the current welding scene to form a track that can reflect the dynamic change trend of the welding working condition.

[0011] Step S111: Extract the working condition change sequence in the historical operation data, refine the working condition conversion logic of different operation stages to form the historical working condition evolution rule, and the historical working condition evolution rule covers the working condition type conversion path and the change trigger condition.

[0012] In this embodiment, the historical operation data includes working condition data, operation behavior data and operation result data of all past welding operations. First, complete welding operation records are selected from the historical operation data, and each record corresponds to a complete welding task. Then, the working condition data of each welding task is arranged in chronological order to form a working condition change sequence, and clustering analysis is performed on multiple working condition change sequences to classify sequences with similar change patterns into a category, and each category corresponds to a typical welding operation scenario. For each typical scenario, the working condition type of different operation stages is analyzed, such as low temperature and low current working condition in preheating stage, high temperature and high current working condition in welding stage, and gradually decreasing temperature and current working condition in cooling stage, and the working condition type conversion path and trigger condition are determined, such as triggering conversion to welding stage when temperature in preheating stage reaches a preset value, triggering conversion to cooling stage when weld length in welding stage reaches a preset value, and finally refining the historical working condition evolution rule including working condition type conversion path and change trigger condition.

[0013] Step S112: The working condition perception components carried by the robot continuously collect real-time working condition data of the current operation scenario, and arrange them in chronological order to form a current real-time working condition flow, and the collection frequency of the current real-time working condition flow is synchronized with the operation behavior execution frequency.

[0014] In this embodiment, the working condition perception components carried by the robot include temperature sensor, current sensor, voltage sensor, weld tracking sensor, etc. The temperature sensor collects welding area temperature data, the current sensor collects welding current data, the voltage sensor collects welding voltage data, and the weld tracking sensor collects weld position, width, depth and other data. These components continuously collect real-time working condition data of the current welding operation scenario at the same frequency as the operation behavior execution frequency, and the collected data is arranged in chronological order to form a current real-time working condition flow. This working condition flow contains all real-time working condition data from the start of the welding operation to the current time, and each data point corresponds to a specific operation time point.

[0015] Step S113: Establish a working condition association mapping mechanism to associate and match the initial working condition in the current real-time working condition flow with the similar initial working condition in the historical working condition evolution rule, and locate the corresponding historical working condition evolution branch.

[0016] In this embodiment, the initial working condition refers to the working condition at the beginning of the welding operation, including the initial temperature of the welding area, the initial welding current, the initial welding voltage, the initial weld position, and other data. First, the initial working condition data is extracted from the current real-time working condition stream, and then the initial working condition data of all typical scenarios is extracted from the historical working condition evolution rule. Then, a working condition association mapping mechanism is established, and the association matching is realized by calculating the similarity between the current initial working condition data and the historical initial working condition data. The similarity calculation is based on the weighted calculation of temperature, current, voltage, and weld position. When the similarity value reaches the preset threshold, it is considered that the current initial working condition is similar to the historical initial working condition, and the corresponding historical working condition evolution branch is located.

[0017] Step S114: Based on the historical working condition evolution branch matched by the association, the subsequent change data of the current real-time working condition stream is combined to supplement the change characteristics specific to the current working condition, and an initial trajectory framework of working condition evolution is constructed.

[0018] In this embodiment, the historical working condition evolution branch matched by the association includes the conversion path and the change trigger condition of the working condition type in this typical scenario. First, the possible working condition evolution path of the current welding operation is predicted according to the historical branch, and then the difference between the current working condition and the corresponding working condition in the historical working condition evolution branch is analyzed by combining the subsequent change data of the current real-time working condition stream. The change characteristics specific to the current working condition are extracted, such as the difference in weld width change rate between the current welding operation and the historical typical scenario, the difference in welding area temperature rise rate between the current welding operation and the historical typical scenario, etc. Finally, these specific change characteristics are supplemented to the predicted working condition evolution path, and the path is corrected and improved to construct an initial trajectory framework that includes the working condition type conversion path and the corresponding change trigger condition from the beginning of the welding operation to the predicted end of the operation.

[0019] Step S115: Continuously correct the trajectory nodes of the initial trajectory framework according to the latest data of the current real-time working condition stream.

[0020] In this embodiment, after the initial trajectory framework is constructed, the robot control system continuously receives the latest data of the current real-time working condition stream. Each time a new real-time working condition data point is received, it is compared with the predicted working condition data at the corresponding operation time point in the initial trajectory framework. If the deviation between the actual working condition data and the predicted working condition data is within the preset range, the initial trajectory framework remains unchanged. If the deviation exceeds the preset range, the corresponding trajectory node in the initial trajectory framework is corrected, including adjusting the trigger condition of working condition type conversion, changing the path of working condition type conversion, etc. By continuously receiving the latest real-time working condition data and correcting the initial trajectory framework, a complete working condition evolution trajectory that accurately reflects the dynamic change trend of the current welding operation working condition with the operation process is finally formed.

[0021] Step S120: generating a job behavior gene segment, extracting an adapted job behavior core element based on the working condition evolution trajectory, combining to form a job behavior gene segment, and the job behavior gene segment carrying basic evolution information of the job action.

[0022] In this embodiment, the generation of the job behavior gene segment is to convert the working condition evolution trajectory into job action information that can be executed by the robot, and the core is to extract the adapted job behavior core element from the working condition evolution trajectory, and combine these elements into a gene segment with basic evolution information.

[0023] Step S121: analyzing the working condition characteristics in each stage of the working condition evolution trajectory, and extracting the core demand of each stage working condition on the job behavior, which is related to the execution mode and connection logic of the job action.

[0024] In this embodiment, the working condition evolution trajectory is first divided according to the job stage, and each job stage corresponds to a continuous trajectory segment in the working condition evolution trajectory, for example, divided into a preheating stage, a welding stage, and a cooling stage. Then, the working condition characteristics of each job stage are analyzed, including temperature gradually increasing, current remaining stable, voltage remaining stable, and weld position fixed in the preheating stage; temperature remaining at a high level, current adjusting according to weld width, voltage adjusting according to current, and weld position gradually changing with the welding process in the welding stage; and temperature gradually decreasing, current gradually decreasing, voltage gradually decreasing, and weld position fixed in the cooling stage. Then, the core demand of each stage on the job behavior is extracted according to the working condition characteristics of each stage, including the robot maintaining a stable welding posture to ensure uniform temperature rise in the welding area in the preheating stage; the robot adjusting the welding posture according to the weld position change, adjusting the welding current and voltage according to the weld width change to ensure the weld quality in the welding stage; and the robot maintaining a stable welding posture until the welding area temperature decreases to a preset value in the cooling stage.

[0025] Step S122: extracting job action units adapted to the core demand of each stage working condition from the historical optimal job behavior database, each job action unit containing independent execution logic and adapted working condition range.

[0026] In this embodiment, the historical optimal job behavior database stores job behavior records verified by practice under different working condition scenarios, and each record contains a complete job action sequence and a corresponding working condition description.

[0027] Step S1221: constructing a historical optimal job behavior database, which stores job behavior records verified by practice under different working condition scenarios, and each job behavior record contains a complete job action sequence and a corresponding working condition description.

[0028] In this embodiment, first, the work behavior records of all past welding operations are collected, each record containing a complete sequence of work actions and corresponding work condition descriptions. The collected records are then screened to select valid records with work results meeting preset quality standards. The valid records are then evaluated, with evaluation indicators including work efficiency, weld quality, energy consumption, etc. The optimal work behavior records are selected according to the evaluation results. Finally, these optimal records are stored according to work condition scenarios to form a historical optimal work behavior database.

[0029] Step S1222: Analyze the work condition characteristics of each stage and convert the work condition characteristics into work demand descriptions, which set the effects to be achieved by the work behavior and the adapted work condition.

[0030] In this embodiment, first, the work condition characteristics of each stage are extracted from the work condition evolution trajectory. The preheating stage work condition characteristics include temperature range, current range, voltage range, and weld position range. The welding stage work condition characteristics include temperature range, current adjustment range, voltage adjustment range, and weld position change range. The cooling stage work condition characteristics include temperature reduction range, current reduction range, voltage reduction range, and weld position fixing range. Then, the work condition characteristics are converted into work demand descriptions. The preheating stage work demand description is that the robot maintains a fixed welding posture, the welding current and voltage are kept within the preset range, and the temperature of the welding area reaches the target temperature of the preheating stage. The welding stage work demand description is that the robot adjusts the welding posture according to the weld position change, adjusts the welding current and voltage according to the weld width change, and the weld length reaches the preset value. The cooling stage work demand description is that the robot maintains a fixed welding posture, the welding current and voltage gradually decrease to the preset value, and the temperature of the welding area decreases to the target temperature of the cooling stage.

[0031] Step S1223: Based on the work demand description, similarity search is performed in the historical optimal work behavior database to screen out work behavior records with matching degrees meeting the requirements.

[0032] In this embodiment, firstly, the job requirement description is converted into retrieval keywords, including job stage, job requirement effect, and adapted working condition, and then the retrieval keywords are used to search in the historical optimal job behavior database. The search method can be keyword-based matching search or semantic-based similarity search. The keyword-based matching search matches the retrieval keywords with the tags of the job behavior records in the database, including job stage, job effect, and working condition. When the number of matching tags reaches a preset number, it is considered that the record is related to the job requirement description. The semantic-based similarity search converts the job requirement description and the working condition description of the job behavior records in the database into vector form, and judges the relevance of the records and the job requirement description by calculating the similarity between the vectors. When the similarity reaches a preset threshold, it is considered that the matching degree of the record and the job requirement description meets the requirements.

[0033] Step S1224: Independent job action units are split from the screened job behavior records. Each job action unit has complete execution logic and can independently respond to specific working condition requirements.

[0034] In this embodiment, the screened job behavior records contain complete job action sequences, for example, the records suitable for the welding stage contain a series of job actions such as adjusting the welding posture, adjusting the welding current, and adjusting the welding voltage. Firstly, the execution logic of each action in the job action sequence is analyzed to determine whether each action can independently complete a specific job task and has complete input and output conditions. For example, the action of adjusting the welding posture can independently complete the task of adjusting the robot welding posture according to the change of the weld position, with the weld position change information as the input condition and the robot welding posture adjustment instruction as the output condition. Therefore, this action can be used as an independent job action unit. The action of adjusting the welding current can independently complete the task of adjusting the welding current according to the change of the weld width, with the weld width change information as the input condition and the welding current adjustment instruction as the output condition. Therefore, this action can also be used as an independent job action unit. Through the analysis of the job action sequence, the actions that can independently respond to specific working condition requirements are split out to form independent job action units.

[0035] Step S1225: The working condition description information corresponding to each job action unit is extracted to define the adaptation working condition range of each job action unit, so that the job action unit is accurately adapted to the core requirement of the corresponding stage.

[0036] In this embodiment, each job action unit comes from a historical optimal job behavior record, and the record contains the working condition description information corresponding to the action unit. First, the working condition description information corresponding to each job action unit is extracted from the job behavior record, for example, the working condition description information corresponding to the job action unit of adjusting the welding posture includes the welding seam position change range, the welding area temperature range, the welding current range, etc.; the working condition description information corresponding to the job action unit of adjusting the welding current includes the welding seam width change range, the welding area temperature range, the welding voltage range, etc. Then, the adaptive working condition range of each job action unit is defined according to the working condition description information, for example, the adaptive working condition range of the job action unit of adjusting the welding posture is that the welding seam position change range is within a certain interval, the welding area temperature range is within a certain interval, and the welding current range is within a certain interval; the adaptive working condition range of the job action unit of adjusting the welding current is that the welding seam width change range is within a certain interval, the welding area temperature range is within a certain interval, and the welding voltage range is within a certain interval.

[0037] Step S123: Extract the core control elements of each job action unit, which include the execution rhythm, intensity correlation characteristics and posture adjustment logic of the job action.

[0038] In this embodiment, different job action units have different core control elements. For the job action unit of adjusting the welding posture, the core control elements include the execution rhythm, intensity correlation characteristics and posture adjustment logic, the execution rhythm refers to the speed of the robot adjusting the welding posture, the intensity correlation characteristics refer to the relationship between the intensity applied by the robot when adjusting the welding posture and the welding seam position change, and the posture adjustment logic refers to the specific way of the robot adjusting the welding posture according to the welding seam position change; for the job action unit of adjusting the welding current, the core control elements include the execution rhythm, intensity correlation characteristics and current adjustment logic, the execution rhythm refers to the speed of the robot adjusting the welding current, the intensity correlation characteristics refer to the relationship between the current value adjusted by the robot when adjusting the welding current and the welding seam width change, and the current adjustment logic refers to the specific way of the robot adjusting the welding current according to the welding seam width change; for the job action unit of adjusting the welding voltage, the core control elements include the execution rhythm, intensity correlation characteristics and voltage adjustment logic, the execution rhythm refers to the speed of the robot adjusting the welding voltage, the intensity correlation characteristics refer to the relationship between the voltage value adjusted by the robot when adjusting the welding voltage and the welding current change, and the voltage adjustment logic refers to the specific way of the robot adjusting the welding voltage according to the welding current change.

[0039] Step S124: Perform feature standardization processing on the core control elements, so that the core control elements of different job action units have a unified representation form.

[0040] In this embodiment, the core control elements of different operation action units can have different representation forms, for example, the execution rhythm of the operation action unit adjusting the welding posture is represented by the number of adjusted angles per second, and the execution rhythm of the operation action unit adjusting the welding current is represented by the current value adjusted per second. First, the standardized range of each core control element is determined, for example, the standardized range of the execution rhythm is 0 to 1, where 0 represents the slowest execution rhythm and 1 represents the fastest execution rhythm; the standardized range of the force correlation feature is 0 to 1, where 0 represents the smallest force correlation and 1 represents the largest force correlation; the logic elements such as posture adjustment logic, current adjustment logic and voltage adjustment logic are converted into standardized numerical forms by encoding. Then, the original value of each core control element is mapped into the standardized range, for example, the original value of the execution rhythm of the operation action unit adjusting the welding posture is the number of adjusted angles per second, which is mapped into the standardized range of 0 to 1 by calculating the difference between the original value of the execution rhythm and the minimum original value of the execution rhythm divided by the difference between the maximum original value and the minimum original value of the execution rhythm.

[0041] Step S125: According to the stage order of the working condition evolution track, the core control elements suitable for each stage working condition are sequentially combined to form operation behavior gene segments, and each operation behavior gene segment corresponds to one stage or multiple continuous stages of the working condition evolution track.

[0042] In this embodiment, the working condition evolution trajectory is divided into a preheating stage, a welding stage, and a cooling stage in sequence. First, determine the adaptive work action unit for each stage. The adaptive work action unit for the preheating stage is the unit that maintains the welding posture. The adaptive work action unit for the welding stage is the unit that adjusts the welding posture, adjusts the welding current, and adjusts the welding voltage. The adaptive work action unit for the cooling stage is the unit that maintains the welding posture and gradually reduces the current and voltage. Then extract the standardized core control elements of each adaptive work action unit. For example, the core control elements of the work action unit that maintains the welding posture in the preheating stage include the execution rhythm standardization value, the intensity correlation feature standardization value, and the posture maintenance logic code value. The core control elements of the work action unit that adjusts the welding posture in the welding stage include the execution rhythm standardization value, the intensity correlation feature standardization value, and the posture adjustment logic code value. The core control elements of the work action unit that adjusts the welding current include the execution rhythm standardization value, the intensity correlation feature standardization value, and the current adjustment logic code value. The core control elements of the work action unit that adjusts the welding voltage include the execution rhythm standardization value, the intensity correlation feature standardization value, and the voltage adjustment logic code value. The core control elements of the work action unit that maintains the welding posture and gradually reduces the current and voltage in the cooling stage include the execution rhythm standardization value, the intensity correlation feature standardization value, the posture maintenance logic code value, the current reduction logic code value, and the voltage reduction logic code value. Finally, combine the core control elements adapted for each stage in sequence according to the stage order of the working condition evolution trajectory. The core control elements of the preheating stage are combined into a work behavior gene segment. The core control elements of the welding stage are combined into a work behavior gene segment. The core control elements of the cooling stage are combined into a work behavior gene segment.

[0043] Step S130: Drive work behavior self-assembly. Based on the work behavior gene segment, combine the real-time change situation of the working condition evolution trajectory to form a dynamic work behavior sequence.

[0044] In this embodiment, work behavior self-assembly is the process of converting work behavior gene segments into dynamic work behavior sequences that can be executed by robots. The core is to combine and adjust work action units based on work behavior gene segments and real-time change situations of working condition evolution trajectories.

[0045] Step S131: Analyze the core control element correlation in the work behavior gene segment, establish gene segment self-assembly rules, and set the combination order and interaction logic of the core control elements.

[0046] In this embodiment, the work behavior gene fragment contains core control elements adapted to different working conditions, and these elements have certain correlations. First, the correlations between the core control elements in the work behavior gene fragment are analyzed. For example, in the work behavior gene fragment of the welding stage, there is a correlation between the core control element for adjusting the welding posture and the core control element for adjusting the welding current. When the welding posture is adjusted, the welding current needs to be adjusted accordingly based on the change in weld width. There is also a correlation between the core control element for adjusting the welding current and the core control element for adjusting the welding voltage. When the welding current is adjusted, the welding voltage needs to be adjusted accordingly based on the change in current. Then, gene fragment self-assembly rules are established based on these correlations. The self-assembly rules set the combination order and interaction logic of the core control elements. For example, the combination order of the core control elements in the welding stage is: first, adjust the welding posture; then, adjust the welding current according to the result of the welding posture adjustment; and finally, adjust the welding voltage according to the result of the welding current adjustment. The interaction logic is: the action unit for adjusting the welding posture sends weld position change information to the action unit for adjusting the welding current, and the action unit for adjusting the welding current sends current change information to the action unit for adjusting the welding voltage.

[0047] Step S132: Monitor the changing trend of the working condition evolution trajectory in real time, identify trajectory nodes that exceed the preset change range in the working condition evolution trajectory, and mark them as working condition change nodes.

[0048] In this embodiment, the changing trends of the operating condition evolution trajectory include the conversion of operating condition types and the rate of change of operating condition parameters. First, a preset range of change is set, which includes the time range for operating condition type conversion and the range of the rate of change of operating condition parameters. For example, the time range for operating condition type conversion during the welding stage is a certain interval, the rate of change of operating condition parameters (temperature, current, etc.) is a certain numerical range per second, and the rate of change of operating condition parameters (current, etc.) is a certain numerical range per second. Then, the real-time monitored changes in the operating condition evolution trajectory are compared with the preset range of change. When the time for operating condition type conversion exceeds the set time range, or the rate of change of operating condition parameters (temperature, current, etc.) exceeds the set rate of change, or the rate of change of operating condition parameters (current, current, etc.) exceeds the set rate of change, the corresponding trajectory node is identified as a node exceeding the preset range of change and marked as an operating condition abrupt change node.

[0049] Step S133: For nodes with sudden changes in working conditions, retrieve the core control elements corresponding to the gene fragments of work behavior, adjust the combination of elements based on self-assembly rules, and generate temporary work action modules that adapt to sudden changes in working conditions.

[0050] In this embodiment, the sudden change node corresponds to the situation where the temperature change rate exceeds the preset range during the welding stage. First, the corresponding core control elements are retrieved from the operational behavior gene fragments of the welding stage. These elements include core control elements for adjusting welding posture, welding current, and welding voltage. Then, the combination of these elements is adjusted based on self-assembly rules. These rules stipulate that when the temperature change rate exceeds the preset range, welding current and voltage should be adjusted first to control the temperature change rate. Therefore, the combination of elements is adjusted by increasing the priority of the core control elements for adjusting welding current and voltage, and decreasing the priority of the core control elements for adjusting welding posture. Specifically, the combination order of the operational action units is changed: first, the welding current is adjusted; then, the welding voltage is adjusted; and finally, the welding posture is adjusted based on the temperature change. Simultaneously, the interaction logic is changed: the action unit for adjusting welding current directly receives information about the temperature change rate and adjusts the current value accordingly; the action unit for adjusting welding voltage directly receives information about the current change and adjusts the voltage value accordingly; and the action unit for adjusting welding posture receives comprehensive information about the temperature change rate and weld position changes and adjusts the posture accordingly.

[0051] Step S134: According to the time sequence of the working condition evolution trajectory, the corresponding work behavior gene fragments of each stage are transformed into basic work action modules, and the basic work action modules are directly adapted to the stage characteristics of the working condition evolution trajectory.

[0052] In this embodiment, the working condition evolution trajectory is divided into a preheating stage, a welding stage, and a cooling stage in chronological order.

[0053] Step S1341: Extract the corresponding work behavior gene fragments for each stage in chronological order of the work condition evolution trajectory.

[0054] In this embodiment, firstly, the operational behavior gene segments corresponding to the preheating stage are extracted in chronological order. The core control element combination in these operational behavior gene segments includes the standardized values ​​of the execution rhythm for maintaining the welding posture, the standardized values ​​of the force-related features, and the posture maintenance logic encoding values. Then, the operational behavior gene segments corresponding to the welding stage are extracted. The core control element combination in these operational behavior gene segments includes the standardized values ​​of the execution rhythm for adjusting the welding posture, the standardized values ​​of the force-related features, and the posture adjustment logic encoding values; the standardized values ​​of the execution rhythm for adjusting the welding current, the standardized values ​​of the force-related features, and the current adjustment logic encoding values; and the standardized values ​​of the execution rhythm for adjusting the welding voltage, the standardized values ​​of the execution rhythm for adjusting the welding voltage, the standardized values ​​of the execution rhythm for adjusting the welding voltage, and the voltage adjustment logic encoding values. Finally, the operational behavior gene segments corresponding to the cooling stage are extracted. The core control element combination in these operational behavior gene segments includes the standardized values ​​of the execution rhythm for maintaining the welding posture, the standardized values ​​of the force-related features, and the posture maintenance logic encoding values; the standardized values ​​of the execution rhythm for gradually decreasing the current, the standardized values ​​of the execution rhythm for gradually decreasing the current, the standardized values ​​of the execution rhythm for gradually decreasing the voltage, the standardized values ​​of the execution rhythm for gradually decreasing the voltage, and the standardized values ​​of the execution rhythm for gradually decreasing the voltage.

[0055] Step S1342: For each job behavior gene fragment, call the job action generation logic to transform the core control elements into a specific set of job action instructions. The execution logic of the set of job action instructions is consistent with the association relationship of the core control elements.

[0056] In this embodiment, the operation action generation logic is a set of rules that transforms standardized core control elements into instructions that the robot can recognize and execute. For the operation behavior gene segment in the preheating stage, the operation action generation logic is invoked to convert the standardized value of the execution rhythm for maintaining the welding posture into a speed command for the robot to maintain the posture, the standardized value of the force association feature into a force command for the robot to maintain the posture, and the posture maintenance logic encoding value into a specific method command for the robot to maintain the posture. For the operation behavior gene segment in the welding stage, the operation action generation logic is invoked to convert the standardized value of the execution rhythm for adjusting the welding posture into a speed command for the robot to adjust the posture, the standardized value of the force association feature into a force command for the robot to adjust the posture, the posture adjustment logic encoding value into a specific method command for the robot to adjust the posture, the standardized value of the execution rhythm for adjusting the welding current into a speed command for the robot to adjust the current, the standardized value of the force association feature into an amplitude command for the robot to adjust the current, the current adjustment logic encoding value into a specific method command for the robot to adjust the current, and the standardized value of the execution rhythm for adjusting the welding voltage into a speed command for the robot to adjust the voltage. The system converts the standardized values ​​of the force-related features into amplitude commands for robot voltage adjustment, and the voltage adjustment logic encoding values ​​into specific method commands for robot voltage adjustment. For the operational behavior gene fragments during the cooling phase, the system calls the operational action generation logic to convert the standardized values ​​of the execution rhythm for maintaining the welding posture into speed commands for robot posture maintenance, the standardized values ​​of the force-related features into force commands for robot posture maintenance, the posture maintenance logic encoding values ​​into specific method commands for robot posture maintenance, the standardized values ​​of the execution rhythm for gradually decreasing the current into speed commands for robot current reduction, the standardized values ​​of the force-related features into amplitude commands for robot current reduction, the current reduction logic encoding values ​​into specific method commands for robot current reduction, the standardized values ​​of the execution rhythm for gradually decreasing the voltage into speed commands for robot voltage reduction, the standardized values ​​of the force-related features into amplitude commands for robot voltage reduction, and the voltage reduction logic encoding values ​​into specific method commands for robot voltage reduction.

[0057] Step S1343: Modularly encapsulate the set of job action instructions to form basic job action modules. Each basic job action module contains a complete action execution process and parameter configuration.

[0058] In this embodiment, modular encapsulation involves dividing and organizing the set of work action instructions according to their functions to form independent modules. For the set of work action instructions in the preheating stage, modular encapsulation is performed according to the function of maintaining the welding posture, forming a basic preheating stage work action module. This module includes the complete execution flow for maintaining the welding posture, including receiving posture-maintaining instructions, adjusting robot joint angles, maintaining the posture until receiving new instructions, and corresponding parameter configurations, including posture-maintaining speed parameters and force parameters. For the set of work action instructions in the welding stage, modular encapsulation is performed according to the functions of adjusting welding posture, adjusting welding current, and adjusting welding voltage, forming a basic welding stage work action module. This module includes the execution flow for adjusting welding posture, adjusting welding current, and adjusting... The execution flow of welding voltage and the corresponding parameter configuration include the speed and force parameters for adjusting the posture, the speed and amplitude parameters for adjusting the current, and the speed and amplitude parameters for adjusting the voltage. For the set of operation instructions in the cooling stage, the functions of maintaining welding posture, gradually reducing current, and gradually reducing voltage are modularly encapsulated to form the basic operation module for the cooling stage. This module includes the execution flow for maintaining welding posture, the execution flow for gradually reducing current, and the execution flow for gradually reducing voltage, as well as the corresponding parameter configuration, including the speed and force parameters for maintaining posture, the speed and amplitude parameters for reducing current, and the speed and amplitude parameters for reducing voltage.

[0059] Step S1344: Associate and bind the basic operation action module with the corresponding working condition evolution trajectory stage features, so that the execution parameters of the basic operation action module accurately match the working condition features of the corresponding working condition evolution trajectory stage.

[0060] In this embodiment, the stage characteristics of the preheating stage's operating condition evolution trajectory are first extracted, including the rate of temperature increase, the range of stable current, and the range of stable voltage. Then, the execution parameters of the basic operation module in the preheating stage are associated with these stage characteristics. For example, the speed parameter for maintaining the welding posture is associated with the rate of temperature increase; when the temperature increase rate is faster, the speed parameter for maintaining the posture is increased accordingly. Next, the stage characteristics of the welding stage's operating condition evolution trajectory are extracted, including the range of high temperature, the range of current adjustment based on weld width, and the range of voltage adjustment based on current. The execution parameters of the basic operation module in the welding stage are associated with these stage characteristics. For example, the speed parameter for adjusting the welding posture is associated with the range of temperature maintenance; when the temperature is maintained within a large range, the speed parameter for adjusting the posture is increased accordingly. Finally, the stage characteristics of the cooling stage's operating condition evolution trajectory are extracted, including the rate of temperature decrease, the range of current decrease, and the range of voltage decrease. The execution parameters of the basic operation module in the cooling stage are associated with these stage characteristics. For example, the speed parameter for maintaining the welding posture is associated with the rate of temperature decrease; when the temperature decrease rate is faster, the speed parameter for maintaining the posture is increased accordingly.

[0061] Step S1345: Analyze the connection logic between basic operation action modules to ensure a smooth transition between basic operation action modules in adjacent stages and avoid interruptions or conflicts during action execution.

[0062] In this embodiment, the basic operation modules for adjacent stages include modules from the preheating stage to the welding stage and modules from the welding stage to the cooling stage. First, the connection logic of the basic operation modules from the preheating stage to the welding stage is analyzed. The preheating stage module maintains the welding posture, while the welding stage module adjusts the welding posture, current, and voltage. The connection logic is as follows: when the temperature in the preheating stage reaches a preset value, the preheating stage basic operation module sends a trigger signal to the welding stage basic operation module. Upon receiving the signal, the welding stage basic operation module first adjusts the welding posture, then adjusts the welding current based on the posture adjustment result, and finally adjusts the welding voltage based on the current adjustment result. During the connection process, it is ensured that the execution parameters of the preheating stage module are consistent with the initial execution parameters of the welding stage module. Smooth transition; Next, we analyze the connection logic of the basic operation action modules from the welding stage to the cooling stage. The module in the welding stage performs actions to adjust the welding posture, current, and voltage. The module in the cooling stage performs actions to maintain the welding posture and gradually reduce the current and voltage. The connection logic is that when the weld length in the welding stage reaches the preset value, the basic operation action module in the welding stage sends a trigger signal to the basic operation action module in the cooling stage. After receiving the signal, the basic operation action module in the cooling stage first performs the action of maintaining the welding posture, then gradually reduces the welding current, and finally gradually reduces the welding voltage. During the connection process, we ensure that the final execution parameters of the welding stage module and the initial execution parameters of the cooling stage module transition smoothly.

[0063] Step S135: Connect and integrate the temporary operation action module and the basic operation action module according to the operation execution logic, eliminate the execution conflict between modules, and form a coherent dynamic operation behavior sequence. The execution rhythm of the dynamic operation behavior sequence is coordinated with the rate of change of the working condition evolution trajectory.

[0064] In this embodiment, the temporary operation action module is generated for nodes with sudden changes in operating conditions, while the basic operation action module is generated for normal operating conditions. First, the job execution logic is determined. This logic specifies the execution order and priority of temporary and basic job action modules. For example, when a sudden change in working condition occurs during the welding stage, the temporary job action module has a higher priority than the basic job action module during the welding stage. The temporary job action module is executed during the sudden change in working condition, and the basic job action module is executed after the working condition returns to normal. Then, the temporary and basic job action modules are integrated and connected according to the job execution logic. For example, during the execution of the basic job action module in the welding stage, when a sudden change in working condition is detected, the execution of the basic job action module is paused, and the execution of the temporary job action module is started. After the sudden change in working condition ends, the execution of the basic job action module is resumed, and the execution result of the temporary job action module is used as the initial parameter for the resumption of the basic job action module. Finally, the execution rhythm of the dynamic job behavior sequence is adjusted to keep in sync with the rate of change of the working condition evolution trajectory. For example, when the rate of change of the working condition evolution trajectory is fast, the execution rhythm of the dynamic job behavior sequence is accelerated accordingly, and when the rate of change of the working condition evolution trajectory is slow, the execution rhythm of the dynamic job behavior sequence is slowed down accordingly.

[0065] Step S140: Capture behavioral condition covariance signals. During the execution of the dynamic work behavior sequence, synchronously capture the covariance information of the changes in work behavior state and the evolution trajectory of the working condition to form behavioral condition covariance signals.

[0066] In this embodiment, the capture of behavioral and working condition covariance signals is to obtain information on the mutual influence between work behavior and working conditions. Its core is to synchronously capture changes in work behavior state and changes in working condition evolution trajectory, and analyze the covariance relationship between the two.

[0067] Step S141: Activate the robot's behavior status monitoring component to collect real-time work behavior status data during the execution of dynamic work behavior sequences. The work behavior status data includes action execution accuracy, posture changes, and the smoothness of action transitions.

[0068] In this embodiment, the robot's behavior state monitoring components include position sensors, angle sensors, and force sensors. Position sensors collect position data from the robot's end effector, angle sensors collect angle data from the robot's joints, and force sensors collect force data applied by the robot during action execution. These components collect real-time work behavior state data during the execution of a dynamic work behavior sequence. Action execution accuracy is measured by comparing the deviation between the actual position of the robot's end effector and a preset position; posture changes are measured by changes in robot joint angles; and the smoothness of action transitions is measured by the time interval between adjacent actions and the smoothness of parameter transitions.

[0069] Step S142: Synchronously collect real-time update data of the working condition evolution trajectory, record the dynamic changes of the working condition during the execution of the work behavior, and the dynamic changes correspond one-to-one with the collection time points of the work behavior status data.

[0070] In this embodiment, the real-time update data of the working condition evolution trajectory includes changes in working condition parameters such as temperature, current, voltage, and weld position. The robot control system synchronously collects this real-time updated data through the working condition sensing component. The time point of collection corresponds one-to-one with the time point of collection of the work behavior status data, ensuring that each work behavior status data point has a corresponding working condition parameter data point.

[0071] Step S143: Establish a covariance information extraction mechanism, compare and analyze the operational behavior status data and dynamic change information of the working conditions at the same time point, and identify the correlation and change patterns between the two.

[0072] In this embodiment, the covariance information extraction mechanism is a set of rules for analyzing the correlation between work behavior status data and dynamic changes in working conditions.

[0073] Step S1431: Perform feature alignment on the work behavior status data and dynamic change information of the working conditions collected at the same time point to make the two types of data comparable in feature dimensions.

[0074] In this embodiment, the feature dimensions of the work behavior status data are first determined, including action execution accuracy, posture change amplitude, and action connection time; the feature dimensions of the dynamic change information of the working condition are also determined, including temperature change amplitude, current change amplitude, voltage change amplitude, and weld position change amplitude. Then, the work behavior status data and dynamic change information of the working condition at the same point in time are aligned in terms of features. For example, the action execution accuracy is aligned with the temperature change amplitude, the posture change amplitude with the weld position change amplitude, and the action connection time with the current change amplitude. The feature alignment method includes normalizing the data to ensure that data of different feature dimensions have the same dimensions and range.

[0075] Step S1432: Construct a two-way correlation analysis model, using the work behavior status data and the dynamic change information of the working conditions as input variables of the model, and conduct a mutual influence analysis.

[0076] In this embodiment, the bidirectional correlation analysis model is used to analyze the degree of mutual influence between two variables. First, a large amount of operational behavior status data and corresponding dynamic change information of working conditions are collected as training data. Then, the training data is preprocessed, including data cleaning and feature extraction. Next, a suitable model structure is selected, such as a neural network model. This neural network model includes an input layer, a hidden layer, and an output layer. The input layer receives the operational behavior status data and the dynamic change information of working conditions, the hidden layer processes the input data, and the output layer outputs the degree of mutual influence. Finally, the model is trained using the training data, and the model parameters are adjusted so that the model can accurately predict the degree of mutual influence between the operational behavior status data and the dynamic change information of working conditions.

[0077] Step S1433: Calculate the trigger weight of the change in work behavior state on the change in working condition, and the driving weight of the change in working condition on the adjustment of work behavior state through the model.

[0078] In this embodiment, the trigger weight refers to the probability and degree to which a change in the work behavior state triggers a change in the working condition, and the driving weight refers to the probability and degree to which a change in the working condition triggers an adjustment in the work behavior state. The bidirectional correlation analysis model can calculate these weight values ​​by learning from training data. For example, when the model inputs data on changes in the work behavior state, it outputs the trigger weight of the change in the work behavior state on the change in the working condition; when the model inputs data on changes in the working condition, it outputs the driving weight of the change in the working condition on the adjustment of the work behavior state.

[0079] Step S1434: Based on the trigger weight and driving weight, identify the positive and negative correlation changes between the operation behavior and the working condition. Positive correlation changes refer to the two showing the same trend, and negative correlation changes refer to the two showing the opposite trend.

[0080] In this embodiment, a positive correlation change refers to a change in the work behavior state and a change in the working condition showing the same trend, while a negative correlation change refers to a change in the work behavior state and a change in the working condition showing opposite trends. Positive and negative correlation changes are identified by comparing the trigger weight and the driving weight with preset thresholds. For example, when the trigger weight of a change in the work behavior state on a change in the working condition is positive and greater than the preset threshold, a positive correlation is considered to exist; when the trigger weight is negative and less than the preset threshold, a negative correlation is considered to exist. Similarly, when the driving weight of a change in the working condition on an adjustment of the work behavior state is positive and greater than the preset threshold, a positive correlation is considered to exist; when the driving weight is negative and less than the preset threshold, a negative correlation is considered to exist.

[0081] Step S1435: Integrate the distribution patterns of positive and negative correlation changes to form a correlation change pattern that can comprehensively reflect the interaction between the two. The correlation change pattern covers the correlation strength and change pattern of different operation stages.

[0082] In this embodiment, the distribution of positive and negative correlation changes in different operation stages is first analyzed. For example, in the preheating stage, positive correlation changes mainly manifest as the correlation between the execution accuracy of the operation and the temperature change amplitude, while negative correlation changes mainly manifest as the correlation between the change amplitude of the operation posture and the current change amplitude. In the welding stage, positive correlation changes mainly manifest as the correlation between the execution accuracy of the operation and the weld position change amplitude, while negative correlation changes mainly manifest as the correlation between the connection time of the operation action and the voltage change amplitude. In the cooling stage, positive correlation changes mainly manifest as the correlation between the execution accuracy of the operation and the temperature change amplitude, while negative correlation changes mainly manifest as the correlation between the change amplitude of the operation posture and the current change amplitude. Then, the correlation change rules are integrated based on the distribution. These rules cover the correlation strength and change pattern in different operation stages. The correlation strength is measured by the magnitude of the trigger weight and the drive weight, and the change pattern is measured by the type of positive and negative correlation changes.

[0083] Step S144: Extract key change features from the correlation change patterns. Key change features can reflect the mutual influence between changes in work behavior status and changes in working conditions.

[0084] In this embodiment, key change features are those features in the correlation change pattern that have important indicative significance for the interaction between work behavior and working conditions. First, all features in the correlation change pattern are evaluated, and the evaluation indicators include the frequency of occurrence of the features and the degree of explanation for the interaction relationship. Then, based on the evaluation results, features with high frequency of occurrence and high degree of explanation are selected as key change features.

[0085] Step S145: Organize the key change features in chronological order to form a continuous behavior and working condition covariance signal. The covariance signal fully records the coordinated changes of work behavior and working conditions during the work process.

[0086] In this embodiment, the key change features are extracted at different time points. These features are organized in chronological order to form a continuous behavioral condition covariance signal. This behavioral condition covariance signal fully records the coordinated changes between the work behavior and the working condition from the start of the welding operation to the current moment, including the key change features at different time points and their temporal relationship.

[0087] Step S150: Iteratively optimize the baseline and behavior scheme, inject the covariance signal of the behavior condition into the generation logic of the condition evolution trajectory, simultaneously optimize the evolution rules of the operation behavior gene fragment, and output the target operation behavior scheme that adapts to the current dynamic condition.

[0088] In this embodiment, iterative optimization is to adjust and optimize the generation logic of the working condition evolution trajectory and the evolution rules of the work behavior gene fragments based on the covariant signal of the working condition. Its core is to achieve continuous improvement of the work behavior scheme by injecting the covariant signal in reverse.

[0089] Step S151: Analyze the key change features in the behavioral operating condition covariance signal and extract optimization indication information that can reflect the trend of operating condition evolution and the adaptation defects of work behavior.

[0090] In this embodiment, the behavioral condition co-variation signal contains key change features at different time points, which reflect the coordinated changes between work behavior and working conditions. First, the key change features in the behavioral condition co-variation signal are analyzed to determine the working condition evolution trend and work behavior adaptation defects indicated by each feature. Then, optimization indication information is extracted based on the analysis results. The optimization indication information includes the working condition evolution trajectory parameters that need to be adjusted and the working behavior gene segment parameters that need to be improved.

[0091] Step S152: Inject the optimization indication information into the generation logic of the working condition evolution trajectory, adjust the fusion weight of historical working condition evolution patterns and current real-time working condition flow, and correct the prediction nodes of the working condition evolution trajectory.

[0092] In this embodiment, the logic for generating the operating condition evolution trajectory includes a fusion rule for historical operating condition evolution patterns and the current real-time operating condition flow. First, optimization indication information is injected into the generation logic. For example, considering the inverse correlation between the timing of work actions and the magnitude of voltage changes, the optimization indication information requires adjusting the prediction parameters for voltage changes. Therefore, this information is injected into the generation logic to adjust the weight of voltage changes in historical operating condition evolution patterns and the weight of voltage data in the current real-time operating condition flow. Then, the fusion weights of historical operating condition evolution patterns and the current real-time operating condition flow are adjusted. The adjustment of the fusion weights is based on the indications of the operating condition evolution trend in the optimization indication information. Next, the prediction nodes of the operating condition evolution trajectory are corrected according to the adjusted fusion weights. The correction of the prediction nodes includes adjusting the predicted time points and the predicted operating condition parameter values.

[0093] Step S153: Based on the revised operating condition evolution trajectory, reassess the adaptability of the job behavior gene fragments and identify the core control elements in the job behavior gene fragments that do not match the revised operating condition evolution trajectory.

[0094] In this embodiment, the fitness assessment of the job behavior gene fragment is the process of determining whether it can accurately respond to the modified job condition evolution trajectory.

[0095] For example, in step S1531: the corrected operating condition evolution trajectory is divided into stages, and the operating condition characteristic parameters and change trends of each stage are extracted.

[0096] In this embodiment, the modified operating condition evolution trajectory is divided into three stages: preheating, welding, and cooling. First, characteristic parameters of the preheating stage are extracted, including the rate of temperature increase, the range of stable current, and the range of stable voltage. The trends of change in the preheating stage are also extracted, including gradually increasing temperature, stable current, and stable voltage. Next, characteristic parameters of the welding stage are extracted, including the range of temperature remaining at a high level, the range of current adjustment based on weld width, the range of voltage adjustment based on current, and the range of weld position changes. The trends of change in the welding stage are also extracted, including maintained high temperature, current adjustment based on weld width, voltage adjustment based on current, and weld position changes with the welding process. Finally, characteristic parameters of the cooling stage are extracted, including the rate of temperature decrease, the range of gradually decreasing current, and the range of gradually decreasing voltage. The trends of change in the cooling stage are also extracted, including gradually decreasing temperature, gradually decreasing current, and gradually decreasing voltage.

[0097] Step S1532: Extract the core control elements corresponding to each stage from the gene fragments of the work behavior, and analyze the degree of fit between the execution logic of the core control elements and the working condition characteristic parameters of the corresponding stage.

[0098] In this embodiment, firstly, the core control elements in the preheating stage operation behavior gene segment are extracted, including the execution rhythm of maintaining welding posture, force correlation characteristics, and posture holding logic, etc., and the degree of adaptation of the execution logic of these core control elements with the characteristic parameters of the preheating stage is analyzed. Next, the core control elements in the welding stage operation behavior gene segment are extracted, including the execution rhythm of adjusting welding posture, force correlation characteristics, and posture adjustment logic; the execution rhythm of adjusting welding current, force correlation characteristics, and current adjustment logic; and the execution rhythm of adjusting welding voltage, force correlation characteristics, and voltage adjustment logic, etc., and the degree of adaptation of the execution logic of these core control elements with the characteristic parameters of the welding stage is analyzed. Finally, the core control elements in the cooling stage operation behavior gene segment are extracted, including the execution rhythm of maintaining welding posture, force correlation characteristics, and posture holding logic; the execution rhythm of gradually decreasing current, force correlation characteristics, and current decreasing logic; and the execution rhythm of gradually decreasing voltage, force correlation characteristics, and voltage decreasing logic, etc., and the degree of adaptation of the execution logic of these core control elements with the characteristic parameters of the cooling stage is analyzed.

[0099] Step S1533: Construct an adaptability evaluation index system to quantitatively evaluate the core control elements from three dimensions: action execution efficiency, working condition response speed, and change adaptability.

[0100] In this embodiment, the adaptability evaluation index system includes indicators across three dimensions: action execution efficiency, operating condition response speed, and change adaptability. The action execution efficiency index measures the efficiency with which the core control element performs its actions; the operating condition response speed index measures the response speed of the core control element to changes in operating conditions; and the change adaptability index measures the adaptability of the core control element to changes in operating conditions. First, quantitative standards are set for each dimension of the index, and then the core control element is quantitatively evaluated based on these standards.

[0101] Step S1534: Based on the evaluation results, select the core control elements whose adaptability does not meet the requirements and mark them as control elements to be optimized.

[0102] In this embodiment, a threshold is first set for the adaptability evaluation index, and then the quantitative evaluation results of the core control elements are compared with the threshold. When the evaluation result of a certain core control element does not reach the threshold, it is considered that its adaptability does not meet the requirements and it is marked as a control element to be optimized.

[0103] Step S154: Optimize the evolution rules of the work behavior gene fragments, adjust the combination and adaptation conditions of the core control elements, so that the work behavior gene fragments can better respond to the modified work condition evolution trajectory.

[0104] In this embodiment, the evolution rules of the job behavior gene fragment include the combination method and adaptation conditions of the core control elements. First, the evolution rules are optimized based on the reasons for the mismatch of the control elements to be optimized. For example, for the control element to be optimized for adjusting the welding posture, the mismatch is due to the unreasonable handling of weld position changes in the execution logic. Therefore, the evolution rules are optimized by adjusting the combination method of the core control elements and adding logic for handling weld position changes. Then, based on the optimized evolution rules, the combination method and adaptation conditions of the core control elements are adjusted. For example, the core control elements for adjusting the welding posture are recombined, and logic for segmented adjustment is added.

[0105] Step S155: Based on the optimized job behavior gene fragments and the corrected job condition evolution trajectory, the target job behavior scheme is reassembled to form a target job behavior scheme that can accurately adapt to the changing needs of the current dynamic job conditions.

[0106] In this embodiment, firstly, based on the corrected working condition evolution trajectory, the appropriate operational behavior gene segments for each stage are determined. For example, in the preheating stage, the optimized operational behavior gene segments for maintaining the welding posture are adapted; in the welding stage, the optimized operational behavior gene segments for adjusting the welding posture, welding current, and welding voltage are adapted; and in the cooling stage, the optimized operational behavior gene segments for maintaining the welding posture and gradually reducing the current and voltage are adapted. Then, these operational behavior gene segments are self-assembled according to the operational execution logic. The self-assembly process includes determining the execution order, interaction logic, and connection method of the operational actions. Finally, the target operational behavior scheme is formed.

[0107] Figure 2 The illustration shows exemplary hardware and software components of an adaptive robot automation system 100 for dynamic working conditions, which can implement the ideas of this application, according to some embodiments of this application. For example, a processor 120 can be used in the adaptive robot automation system 100 for dynamic working conditions and to perform the functions in this application.

[0108] The adaptive system 100 for robot automation operations under dynamic working conditions can be a general-purpose server or a special-purpose server; both can be used to implement the adaptive method for robot automation operations under dynamic working conditions described in this application. Although only one server is shown in this application, for convenience, the functions described in this application can be implemented in a distributed manner on multiple similar platforms to balance the load.

[0109] For example, the adaptive robot automation system 100 for dynamic working conditions may include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and various forms of storage media 140, such as a disk, ROM, or RAM, or any combination thereof. Exemplarily, the adaptive robot automation system 100 for dynamic working conditions may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The methods of this application can be implemented according to these program instructions. The adaptive robot automation system 100 for dynamic working conditions also includes an I / O interface 150 between the computer and other input / output devices.

[0110] For ease of explanation, only one processor is described in the adaptive robot automation system 100 for dynamic working conditions. However, it should be noted that the adaptive robot automation system 100 for dynamic working conditions may also include multiple processors, and therefore the steps performed by one processor as described in this application may also be performed jointly or individually by multiple processors. For example, if the processor of the adaptive robot automation system 100 for dynamic working conditions performs steps A and B, it should be understood that steps A and B may also be performed jointly by two different processors or individually by one processor. For example, the first processor performs step A, the second processor performs step B, or the first processor and the second processor jointly perform steps A and B.

[0111] Furthermore, this embodiment of the invention also provides a readable storage medium, wherein computer-executable instructions are preset in the readable storage medium, and when the processor executes the computer-executable instructions, the above-mentioned adaptive method for robot automated operation under dynamic working conditions is implemented.

[0112] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. An adaptive method for automated robot operation under dynamic working conditions, characterized in that, The method includes: Anchoring the working condition evolution trajectory, the historical working condition evolution pattern is integrated with the current real-time working condition flow to generate the working condition evolution trajectory. The working condition evolution trajectory represents the dynamic change trend of the working condition as the operation progresses. Gene fragments of work behavior are generated, and core elements of work behavior that are adapted are extracted based on the work condition evolution trajectory. These elements are then combined to form work behavior gene fragments, which carry the basic evolutionary information of work actions. Driven by the self-assembly of work behaviors, based on work behavior gene fragments and combined with the real-time changes in the trajectory of work condition evolution, a dynamic work behavior sequence is formed through self-assembly. Capture covariant signals of behavioral conditions. During the execution of dynamic work behavior sequences, synchronously capture covariant information of changes in work behavior state and work condition evolution trajectory to form covariant signals of behavioral conditions. Iteratively optimize the baseline and behavior scheme, inject the covariance signal of behavior conditions into the generation logic of the work condition evolution trajectory, simultaneously optimize the evolution rules of the work behavior gene fragments, and output the target work behavior scheme that adapts to the current dynamic work conditions.

2. The adaptive method for automated robot operation under dynamic working conditions according to claim 1, characterized in that, The anchored operating condition evolution trajectory, which integrates historical operating condition evolution patterns with the current real-time operating condition flow to generate the operating condition evolution trajectory, includes: Extract the sequence of working conditions changes from historical operation data, refine the working condition transition logic of different operation stages to form historical working condition evolution rules, and the historical working condition evolution rules cover the working condition type transition path and change triggering conditions. The robot continuously collects real-time working condition data of the current work scenario through the working condition perception component, and organizes it into a real-time working condition stream according to the working time sequence. The collection frequency of the real-time working condition stream is synchronized with the execution frequency of the work behavior. Establish a working condition association mapping mechanism to associate and match the initial working condition in the current real-time working condition stream with similar initial working conditions in the historical working condition evolution pattern, and locate the corresponding historical working condition evolution branch. Based on the historical operating condition evolution branches obtained through correlation matching, and combined with the subsequent change data of the current real-time operating condition flow, the unique change characteristics of the current operating condition are supplemented to construct the initial trajectory framework of the operating condition evolution. The trajectory nodes of the initial trajectory framework are continuously revised based on the latest data from the current real-time operating condition stream.

3. The adaptive method for automated robot operation under dynamic working conditions according to claim 1, characterized in that, The generated work behavior gene fragments are formed by extracting suitable core elements of work behavior based on the work condition evolution trajectory and combining them, including: Analyze the characteristics of each stage of the work condition evolution trajectory, extract the core requirements of each stage of the work condition for the work behavior, and the core requirements are related to the execution method and connection logic of the work actions; Extract work action units that are adapted to the core requirements of each stage of work from the historical best work behavior database. Each work action unit contains independent execution logic and adapted work condition range. Extract the core control elements of each work action unit. The core control elements cover the execution rhythm, force correlation characteristics and posture adjustment logic of the work action. The core control elements are standardized to give them a unified representation. According to the stage sequence of the working condition evolution trajectory, the core control elements adapted to each stage of the working condition are combined in an orderly manner to form working behavior gene segments. Each working behavior gene segment corresponds to one or more consecutive stages of the working condition evolution trajectory.

4. The adaptive method for automated robot operation under dynamic working conditions according to claim 1, characterized in that, The self-assembly of driving work behaviors, based on work behavior gene fragments and combined with the real-time changes in the work condition evolution trajectory, self-assembles into a dynamic work behavior sequence, including: Analyze the relationships between core control elements in the gene fragments of work behavior, establish gene fragment self-assembly rules, and set the combination order and interaction logic of core control elements according to the self-assembly rules. Real-time monitoring of the changing trends of the operating condition evolution trajectory, identification of trajectory nodes that exceed the preset range of change in the operating condition evolution trajectory, and marking them as operating condition abrupt change nodes; For nodes with sudden changes in working conditions, the core control elements corresponding to the gene segments of work behavior are retrieved, and the combination of elements is adjusted based on self-assembly rules to generate temporary work action modules that adapt to sudden working conditions. According to the time sequence of the working condition evolution trajectory, the corresponding work behavior gene fragments of each stage are transformed into basic work action modules, and the basic work action modules are directly adapted to the stage characteristics of the working condition evolution trajectory. Temporary operation action modules and basic operation action modules are connected and integrated according to the operation execution logic to eliminate execution conflicts between modules and form a coherent dynamic operation behavior sequence. The execution rhythm of the dynamic operation behavior sequence is coordinated with the rate of change of the working condition evolution trajectory.

5. The adaptive method for automated robot operation under dynamic working conditions according to claim 3, characterized in that, The process involves extracting work action units from a historical best-performing work behavior database that are adapted to the core requirements of each stage of the work. Each work action unit contains independent execution logic and an adapted work condition range, including: Construct a historical best operation behavior database. The database stores operation behavior records that have been verified in practice under different working conditions. Each operation behavior record contains a complete sequence of operation actions and a corresponding working condition description. Analyze the characteristics of the working conditions at each stage, transform these characteristics into job requirement descriptions, and define the effects that the job behavior needs to achieve and the appropriate working conditions. Based on the job requirement description, a similarity search is performed in the historical best job behavior database to filter out job behavior records that meet the requirements of matching the job requirement description. Independent work action units are extracted from the selected work behavior records. Each work action unit has complete execution logic and can independently respond to specific work condition requirements. Extract the working condition description information corresponding to each work action unit, define the applicable working condition range of each work action unit, and make the work action unit accurately match the core working condition requirements of the corresponding stage.

6. The adaptive method for automated robot operation under dynamic working conditions according to claim 4, characterized in that, The process involves converting the corresponding operational behavior gene fragments for each stage into basic operational action modules according to the time sequence of the operational condition evolution trajectory. These basic operational action modules are directly adapted to the stage characteristics of the operational condition evolution trajectory, including: According to the time sequence of the work condition evolution trajectory, extract the corresponding work behavior gene fragments for each stage in turn; For each job behavior gene segment, the job action generation logic is invoked to transform the core control elements into a specific set of job action instructions. The execution logic of the job action instruction set is consistent with the association relationship of the core control elements. The set of job action instructions is modularly encapsulated to form basic job action modules. Each basic job action module contains a complete action execution process and parameter configuration. The basic operation action module is associated and bound with the corresponding work condition evolution trajectory stage characteristics, so that the execution parameters of the basic operation action module accurately match the work condition characteristics of the corresponding work condition evolution trajectory stage. Analyze the connection logic between basic operation action modules to ensure a smooth transition between basic operation action modules in adjacent stages and avoid interruptions or conflicts during action execution.

7. The adaptive method for automated robot operation under dynamic working conditions according to claim 1, characterized in that, The captured behavioral condition covariance signal, during the execution of a dynamic work behavior sequence, synchronously captures the covariance information of work behavior state changes and work condition evolution trajectories to form a behavioral condition covariance signal, including: Activate the robot's behavior status monitoring component to collect real-time work behavior status data during the execution of dynamic work behavior sequences. The work behavior status data includes action execution accuracy, posture changes, and the smoothness of action transitions. Synchronously collect real-time updated data of the working condition evolution trajectory, record the dynamic changes of the working condition during the execution of the work behavior, and the dynamic changes correspond one-to-one with the collection time points of the work behavior status data. Establish a covariance information extraction mechanism, compare and analyze the operational behavior status data and dynamic changes in working conditions at the same point in time, and identify the correlation and change patterns between the two. Extract key change features from the correlation change patterns. These key change features can reflect the mutual influence between changes in work behavior status and changes in working conditions. Key change characteristics are organized in chronological order to form continuous behavioral condition co-variation signals. These co-variation signals fully record the coordinated changes in work behavior and conditions during the work process.

8. The adaptive method for automated robot operation under dynamic working conditions according to claim 7, characterized in that, The establishment of a covariance information extraction mechanism, which compares and analyzes operational behavior status data and dynamic changes in operating conditions at the same point in time, and identifies the correlation patterns between the two, includes: Feature alignment is performed on the operational behavior status data and dynamic change information of the working conditions collected at the same time point to make the two types of data comparable in feature dimensions. A two-way correlation analysis model was constructed, with the operational behavior status data and dynamic changes in working conditions as input variables, and the degree of mutual influence was analyzed. The model calculates the triggering weight of changes in work behavior state on changes in work conditions, and the driving weight of changes in work conditions on adjustments to work behavior state. Based on trigger weights and driving weights, we identify positive and negative correlations between work behaviors and working conditions. Positive correlations refer to the two showing the same trend, while negative correlations refer to the two showing the opposite trend. By integrating the distribution patterns of positive and negative correlation changes, a correlation change pattern that can comprehensively reflect the interaction between the two is formed, covering the correlation strength and change pattern at different operational stages.

9. The adaptive method for automated robot operation under dynamic working conditions according to claim 1, characterized in that, The iterative optimization baseline and behavior scheme injects the co-variant signal of the behavior condition into the generation logic of the condition evolution trajectory, synchronously optimizes the evolution rules of the work behavior gene fragments, and outputs a target work behavior scheme adapted to the current dynamic working condition, including: Analyze the key changing features in the covariant signals of behavioral operating conditions, and extract optimization indication information that can reflect the evolution trend of operating conditions and the adaptation defects of work behavior; The optimization indication information is injected into the generation logic of the working condition evolution trajectory, the fusion weight of historical working condition evolution patterns and current real-time working condition flow is adjusted, and the prediction nodes of the working condition evolution trajectory are corrected. Based on the revised operating condition evolution trajectory, the adaptability of the job behavior gene fragments is reassessed, and the core control elements in the job behavior gene fragments that do not match the revised operating condition evolution trajectory are identified. Optimize the evolution rules of the work behavior gene fragments, adjust the combination and adaptation conditions of the core control elements, so that the work behavior gene fragments can better respond to the modified work condition evolution trajectory. Based on the optimized job behavior gene fragments and the corrected job condition evolution trajectory, the target job behavior scheme is reassembled and formed, which can accurately adapt to the changing needs of the current dynamic job conditions.

10. An adaptive system for automated robot operation under dynamic working conditions, characterized in that, The adaptive system for robot automated operation under dynamic working conditions includes a processor and a memory, the memory and the processor being connected. The memory is used to store programs, instructions or code, and the processor is used to execute the programs, instructions or code in the memory to implement the adaptive method for robot automated operation under dynamic working conditions as described in any one of claims 1-9.

Images