A control method for a crankshaft clamping machine

Abstract

The application discloses a control method of a crankshaft clamping machine, and relates to the technical field of automatic assembly.The clamping machine comprises a truss, a cross beam, a horizontal servo sliding table, a lifting rod, a rotary cylinder, a butt joint seat, two groups of pneumatic clamping claws and a logic controller.The technical key points are as follows: from the equipment initialization self-checking, single workpiece exclusive control parameter pre-generation, to the crankshaft alignment grabbing, the synchronous adjustment of the transfer posture, the hierarchical anti-interference assembly placement, the complete process of the cycle reset and the logic iteration, all are automatically planned and executed by the logic controller, without the need for manual full-process participation in the alignment, hoisting, posture adjustment and other operations, which not only directly reduces the enterprise labor cost, but also completely eliminates the occupational fatigue and labor intensity problems caused by the manual repetitive operation, and fully meets the development requirements of lean production.

Description

Technical Field

[0001] This invention relates to the field of automated assembly technology, and in particular to a control method for a crankshaft clamping machine. Background Technology

[0002] The crankshaft is the core moving component of an engine, undertaking the core function of converting the reciprocating motion of the piston into rotational torque. Its assembly positioning accuracy, transfer efficiency, and operational stability directly determine the assembly quality, operating performance, service life, and production cycle of the final assembly line of the engine. As the internal combustion engine manufacturing industry rapidly develops towards high automation, high cycle time, high precision, and flexibility, the engine model iteration speed is accelerating, which places increasingly higher demands on the automation level, positioning accuracy, operational reliability, model compatibility, and full-process quality control capabilities of the crankshaft transfer and assembly process.

[0003] Currently, in the crankshaft assembly process of domestic engine final assembly lines, the mainstream operation mode of "tooling cart + AGV transfer + manual hoisting" is widely adopted in the industry. Taking the crankshaft final assembly process of the widely used S04 engine model as an example, after the gear pressing process is completed, the operator places the processed and qualified crankshaft smoothly on a special crankshaft tooling cart, completing the material preparation for loading a single crankshaft; the tooling cart carrying the crankshaft is then transferred to the crankshaft assembly station on the final assembly line via an AGV cart along a preset route in the workshop, completing the assembly process. The crankshaft is transferred across processes. Dedicated operators at the assembly station use a cantilever crane to lift the crankshaft by securing the main journals at both ends with a special lifting tool. After lifting the crankshaft, it is manually aligned with the assembly reference of the engine block by visual inspection. The crankshaft is then manually fine-tuned and slowly lowered into the designated assembly position on the block. Once the operator confirms that the crankshaft is in place without interference or misalignment, the lifting tool is removed, completing the assembly of a single crankshaft. The unloaded tooling vehicle is then transferred back to the previous process by an AGV to begin the next work cycle.

[0004] In the actual implementation of the above-mentioned existing technical solutions, two major unsolvable core technical defects were discovered:

[0005] Firstly, under the existing operating model, the cumulative time for a single cycle of crankshaft transfer and assembly exceeds the 80-second design cycle time requirement of the entire engine assembly line, becoming a core bottleneck restricting the overall operating efficiency of the assembly line. To maintain production continuity, four dedicated crankshaft tooling vehicles and one spare AGV are required on-site, resulting in high equipment occupancy and supporting investment costs. Simultaneously, due to workshop layout limitations, the AGV transfer route is circuitous and lengthy, with insufficient inherent transfer efficiency, necessitating frequent manual assistance via hydraulic trucks, further increasing process redundancy and failing to meet the demands of large-volume, high-cycle automated continuous production. Furthermore, throughout the entire operation, crankshaft hoisting, alignment, attitude fine-tuning, and transfer assistance all require manual participation, necessitating two dedicated operators per production line. This not only directly increases the company's labor costs, but the long-term repetitive hoisting operations also easily lead to operator fatigue and high labor intensity, which does not meet the requirements of lean production development.

[0006] Secondly, both manual hoisting and simple truss equipment use open-loop control logic with fixed parameters. This makes it impossible to adapt to individual differences in the machining dimensions, form and position tolerances, and material properties of a single crankshaft. Furthermore, it cannot adaptively compensate for changes in working conditions such as fluctuations in workshop temperature and humidity, air pressure fluctuations, and cumulative equipment wear. Long-term operation can easily lead to problems such as insufficient clamping force, decreased positioning accuracy, and attitude adjustment deviations. This can result in quality issues such as crankshaft journal scratches, loosening and misalignment of bearings, and excessive assembly coaxiality. In severe cases, it can even cause safety accidents such as crankshaft falls and equipment collisions, compromising operational stability. The existing solutions cannot meet the requirements for long-term continuous production. At the same time, they lack the ability to collect and control data throughout the entire process. The data on crankshaft pre-processing, assembly process, subsequent whole machine test and vehicle service are completely disconnected. When quality problems occur, it is impossible to quickly locate the root cause, and it is also impossible to optimize the assembly process based on the whole machine service performance. In addition, there are no tamper-proof records of the assembly process, which cannot meet the automotive industry's compliance control requirements for full life cycle traceability of core components and does not conform to the development trend of digitalization and intelligence in modern manufacturing. Therefore, we propose a control method for crankshaft clamping machines. Summary of the Invention

[0007] (a) Technical problems to be solved

[0008] To address the shortcomings of existing technologies, this invention provides a crankshaft clamping machine, which solves the technical problems of poor stability and inadequate fixing effect of existing crankshaft clamping equipment during use.

[0009] (II) Technical Solution

[0010] To achieve the above objectives, the present invention provides the following technical solution:

[0011] A crankshaft clamping machine includes a truss and two sets of pneumatic grippers for gripping the crankshaft. A crossbeam is mounted on the top of the truss, and the crossbeam has a horizontal servo slide that can reciprocate horizontally along its length. A lifting rod is mounted on the horizontal servo slide, which can move horizontally synchronously with the horizontal servo slide and reciprocate vertically. The two sets of pneumatic grippers are respectively located on both sides of the bottom of the lifting rod and connected to a rotary cylinder at the bottom of the lifting rod. The crankshaft clamping machine also includes a logic controller, which is electrically connected to the horizontal servo slide, the lifting rod, the rotary cylinder, and the pneumatic grippers, and is used to send control commands and receive operating status feedback signals. The logic controller establishes bidirectional encrypted data communication with a supporting edge computing node and a supporting upstream and downstream management and control system of the industrial chain. The control method includes the following steps:

[0012] S1. Return each execution component to its preset initial position. After the device performs a full-dimensional self-check and all components are confirmed to be in place, it enters the standby state.

[0013] S2. Obtain the full-dimensional correlation data of the single crankshaft to be assembled, the matching engine block, and the corresponding whole machine. After transmitting the data to the edge computing node for processing, generate a personalized control parameter group that is uniquely bound to the single crankshaft.

[0014] S3. After receiving the trigger signal that the crankshaft is in place, control the horizontal servo slide and lifting rod according to the personalized control parameter group to complete the gripping and alignment, and then control the pneumatic gripper to complete the adaptive gripping of the crankshaft. After the gripping state meets the preset requirements, control the lifting rod to drive the crankshaft to rise to a safe height and complete the gripping action.

[0015] S4. According to the personalized control parameter group, control the horizontal servo slide to drive the clamped crankshaft to the assembly station for transfer. During the transfer, control the rotary cylinder to dynamically adjust the spatial attitude of the crankshaft and correct the attitude deviation so that when the crankshaft reaches the top of the assembly station, the attitude matches the cylinder mounting reference.

[0016] S5. After receiving the trigger signal that the cylinder block is in place, control the lifting rod to drive the crankshaft to descend in stages according to the personalized control parameter group, so as to complete the stable placement of the crankshaft in the designated position of the engine cylinder block. After the placement is in place, control the pneumatic gripper to release the crankshaft, and the lifting rod to rise to a safe height to complete the assembly action.

[0017] S6. Control all execution components to return to their initial positions, complete the encrypted archiving of the entire process data of this operation, and simultaneously perform iterative optimization of the control logic based on the actual operation data of this operation, waiting for the next operation trigger signal.

[0018] Preferably, in step S1, the device's full-dimensional self-test specifically includes:

[0019] The logic controller performs item-by-item detection on the equipment's air supply pressure, the operating status of each actuator, the communication link status of each component, the communication status of the upstream and downstream supporting control systems in the industrial chain, and the operating status of the edge computing nodes.

[0020] When all test results meet the preset standards and a feedback signal indicating that all actuators have returned to their initial positions is received, the equipment enters the standby state; if the self-test fails, a fault alarm for the corresponding item is triggered and the automatic operation permission is locked.

[0021] Preferably, in step S2, the full-dimensional correlation data of the single crankshaft to be assembled, the matching engine block, and the corresponding whole machine includes the full-process machining dimension data, form and position tolerance data, dynamic balance test data, and material performance test data of the single crankshaft; the machining dimension data, assembly tolerance data, and bearing performance parameters of the matching engine block; and the design performance indicators and subsequent assembly requirements of the corresponding whole machine.

[0022] The edge computing node constructs a dedicated assembly adaptation control model for the single crankshaft using the aforementioned multi-dimensional associated data. Through simulation calculations of the assembly adaptation control model, it generates the personalized control parameter set, which includes the target clamping journal position, target clamping force control curve, gripping alignment coordinates, assembly alignment coordinates, target rotation angle control curve, graded lifting stroke, graded running speed curve, and clamping and holding pressure duration.

[0023] Preferably, in step S2, when the logic controller constructs the assembly adaptation control model for a single crankshaft, it simultaneously acquires the current ambient temperature, ambient humidity, air source temperature, air source dew point, grid voltage fluctuation, and cumulative equipment running time data through the workshop sensing system. The above data is input into the assembly adaptation control model to perform multi-dimensional dynamic correction and compensation on the personalized control parameter group, thereby offsetting the impact of environmental variables and equipment wear on assembly accuracy.

[0024] Throughout the entire operation, the logic controller acquires the action sequence plan and real-time air consumption data of all pneumatic equipment in the production line, predicts the air source pressure fluctuation trend and peak air consumption range, and adjusts the execution sequence window of the pneumatic action of the equipment under the premise of meeting the production line cycle requirements, while dynamically adjusting the output pressure compensation coefficient of the pneumatic components.

[0025] Preferably, step S3 specifically involves: after the logic controller receives the crankshaft loading trigger signal sent by the crankshaft conveying device, it controls the horizontal servo slide to move the lifting rod horizontally to the gripping position directly above the crankshaft target clamping journal according to the personalized control parameter group, and then controls the lifting rod to descend to the preset gripping height according to the graded lifting stroke.

[0026] Then, the clamping extension arms of the pneumatic gripper are controlled to close in opposite directions, and the clamping force is dynamically adjusted according to the target clamping force control curve in the personalized control parameter group to clamp the target clamping journal position of the crankshaft.

[0027] Upon receiving dual feedback signals indicating that the clamping position is in place and the clamping force deviates from the preset value within the threshold range, the control lifting rod drives the clamped crankshaft to rise to the preset safe height, completing the gripping action.

[0028] Preferably, step S4 specifically involves the logic controller controlling the horizontal servo slide to move the clamped crankshaft horizontally to the assembly station according to the personalized control parameter group.

[0029] Throughout the horizontal movement, the synchronously controlled rotary cylinder adjusts the spatial attitude of the crankshaft according to the target rotation angle control curve within the personalized control parameter group, and corrects the crankshaft's runout and torsional errors in real time.

[0030] The attitude adjustment action stops only after the crankshaft reaches directly above the assembly station and a feedback signal is received indicating that the crankshaft attitude matches the preset requirements for the cylinder block mounting reference.

[0031] Preferably, step S5 specifically involves: after the logic controller receives the cylinder block arrival trigger signal sent by the cylinder block conveying device, it performs a full-process pre-simulation of the crankshaft lowering assembly process through the assembly adaptation control model, checks for assembly interference risks, and corrects the staged lifting stroke and running speed curves.

[0032] According to the revised personalized control parameter group, the control lift rod drives the crankshaft to descend in stages. During the descent, the distance data between the lower end face of the crankshaft and the cylinder block mounting surface is collected in real time. When the distance is greater than the preset safety threshold, the crankshaft descends at the first speed. When the distance is less than or equal to the preset safety threshold, the crankshaft descends at the second speed. The second speed is less than the first speed until the crankshaft is stably placed in the designated assembly position of the cylinder block.

[0033] Only after receiving a feedback signal that the crankshaft has been placed in place will the pneumatic gripper's clamping extension arm be opened to release the crankshaft, and then the lifting rod be raised back to the preset safe height to complete the assembly action.

[0034] Preferably, step S6 specifically involves: the logic controller controlling the rotary cylinder, horizontal servo slide, and lifting rod to return to their initial positions in sequence; simultaneously inputting the actual operation data of the entire process of this operation and the assembly status detection data into the assembly adaptation control model to perform virtual-real calibration and iterative optimization of the model;

[0035] After encrypting the calibrated assembly adaptation control model, personalized control parameter group, and full-process operation data, they are bound to the unique identification number of the corresponding crankshaft and cylinder block, and simultaneously uploaded to the upstream and downstream supporting management and control system of the industry chain for archiving.

[0036] At the same time, based on the data from this operation, the assembly adaptation control model construction rules for crankshafts of the same model are optimized, and the operation trigger signal for the next crankshaft is awaited.

[0037] Preferably, the logic controller establishes bidirectional data communication with the crankshaft conveying device, cylinder block conveying device, and upstream and downstream production equipment in the production line through a time-sensitive network. All connected devices use a unified clock source to generate microsecond-level timestamps. Based on the timestamped operating data, the controller performs microsecond-level timing synchronization control across the entire production line, so that the crankshaft clamping process is matched with the actions of upstream and downstream equipment on the production line.

[0038] Meanwhile, the logic controller establishes distributed cluster communication with all crankshaft clamping machines of the same model in the workshop through edge computing nodes, and obtains the cumulative runtime, real-time load rate, equipment health status and number of tasks to be executed for each device in the cluster in real time. It dynamically allocates clamping tasks to be executed through a collaborative scheduling algorithm, keeping the load rate and cumulative runtime of each device in the cluster within a preset balance range.

[0039] When archiving a job, the logic controller uploads the encrypted full-process job data to the consortium blockchain platform for tamper-proof notarization, generating a notarization hash value that is strongly bound to the unique identification number of the corresponding crankshaft, cylinder block, and whole machine.

[0040] Preferably, during the entire operation process from steps S1 to S6, the logic controller performs dual-channel distributed mirror storage of the equipment's full-state operation data at a preset fixed sampling frequency. The full-state operation data includes real-time position data of each component, operation status data, clamping status data, crankshaft unique identification number, executed step data, unexecuted step data, and real-time communication data.

[0041] When the equipment experiences an abnormal power outage or communication interruption, the logic controller uses the backup power supply to encrypt and store the current status data. After the equipment returns to normal, it automatically verifies the integrity of the dual-channel stored data. Once confirmed, it continues to complete the remaining operation steps from the execution breakpoint where the abnormal operation occurred.

[0042] Meanwhile, the logic controller collects the operating electrical signals, vibration signals, action timing signals, and force feedback signals of each actuator in real time, extracts signal feature values ​​and inputs them into the pre-trained fault prediction model to identify early potential fault risks of the components, adjusts the operating parameters of the corresponding components to reduce the load, and triggers the generation of fault warnings and maintenance guidelines and uploads them to the enterprise equipment management system.

[0043] (III) Beneficial Effects

[0044] 1. By employing pneumatic grippers as the crankshaft clamping mechanism, the connecting rod journals on the crankshaft are gripped. Then, the servo slide on the crossbar and the lifting rod on the servo slide drive the crankshaft to move horizontally and vertically, thereby moving the crankshaft to the designated position. Next, the crankshaft's orientation is adjusted by the rotating cylinder at the bottom of the lifting rod, so that the crankshaft mates with the cylinder block. Therefore, the technical problem of insufficient automation in the use of existing crankshaft assembly equipment is effectively solved, thereby reducing the equipment occupation and supporting investment in the production line, while significantly improving the overall operating efficiency and continuous production capacity of the engine assembly line.

[0045] 2. From equipment initialization self-test and pre-generation of dedicated control parameters for each workpiece, to crankshaft alignment and gripping, synchronous adjustment of transport posture, hierarchical anti-interference assembly and placement, cyclic reset, and logical iteration, the entire process is automatically executed by the logic controller. No manual intervention is required for alignment, hoisting, and posture adjustment, directly reducing labor costs and eliminating occupational fatigue and labor intensity caused by repetitive manual work, fully meeting the requirements of lean manufacturing. Secondly, by constructing and implementing a dedicated assembly adaptation control model for each crankshaft, comprehensive data on crankshaft processing and material properties, cylinder block assembly parameters, and overall machine design requirements are pre-acquired, generating a unique set of personalized control parameters for each crankshaft. Throughout the alignment, gripping, transport posture adjustment, and hierarchical assembly and placement process, closed-loop control is executed according to this dedicated parameter set, accurately adapting to individual differences in crankshaft processing dimensions, geometric tolerances, and material properties, while also achieving dynamic compensation for environmental fluctuations, changes in air pressure, and cumulative equipment wear.

[0046] 3. Through the implementation of the full-condition adaptive compensation scheme, during the assembly adaptation control model construction stage, all-dimensional operating condition parameters such as workshop environmental temperature and humidity, air source conditions, power grid voltage fluctuations, and cumulative equipment operating time are simultaneously incorporated to pre-correct and compensate for personalized control parameter groups. Throughout the operation, the action plans and air consumption data of pneumatic equipment across the entire production line are acquired in real time, the trend of air source pressure fluctuations and peak air consumption intervals are predicted, and the timing of pneumatic actions is proactively adjusted without affecting the production line cycle time, while simultaneously and dynamically adjusting the pressure compensation coefficient of pneumatic components. Secondly, through the implementation of the dual-path distributed mirror storage and breakpoint resume solution, the equipment's full-state operating data is backed up in real time at a fixed frequency throughout the operation. When abnormal operating conditions such as abnormal power outages or communication interruptions occur, the current state data can be encrypted and stored through the backup power supply. After the equipment recovers, the remaining work can be resumed directly from the breakpoint without the need for a full-process reset and rework. At the same time, through the real-time acquisition of multi-source operating data and the application of fault prediction models, early potential fault risks of equipment components can be identified, operating parameters can be automatically adjusted to reduce load, and early warning maintenance can be triggered.

[0047] 4. By implementing a time-sensitive network communication scheme, equipment can achieve microsecond-level time synchronization control with upstream and downstream conveyor equipment and process production equipment on the production line. This completely eliminates the transmission of hardware I / O trigger signals, removing delay errors and signal interference caused by hardware triggers. It enables zero-wait collaborative operation across the entire production line, effectively avoiding workstation accumulation and material shortage problems. At the same time, by establishing distributed cluster communication with the same type of equipment in the workshop through edge computing nodes, and dynamically allocating tasks to be executed through collaborative scheduling algorithms, the load rate and cumulative running time of multiple devices are kept within a balanced range. When a single device fails, the task can be automatically taken over. Attached Figure Description

[0048] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0049] Figure 1 This is an overall structural diagram of an embodiment of the present invention;

[0050] Figure 2 This is an assembly diagram of the clamping mechanism in an embodiment of the present invention;

[0051] Figure 3 This is an exploded view of the clamping mechanism in an embodiment of the present invention;

[0052] Figure 4 This is a structural diagram of the pneumatic gripper in an embodiment of the present invention;

[0053] Figure 5 This is a schematic diagram of the docking between the pneumatic gripper and the crankshaft in an embodiment of the present invention;

[0054] Figure 6 This is a core flowchart of the crankshaft clamping machine control method in an embodiment of the present invention.

[0055] Legend:

[0056] 11. Column; 12. Horizontal beam; 13. Lifting rod; 14. Rotary cylinder;

[0057] 2. Cylinder block;

[0058] 3. Clamping mechanism; 31. Docking seat; 32. Clamping component; 321. Pneumatic gripper; 322. Fixed clamp; 323. Movable gripper; 324. Extending gripper; 325. Flexible pad;

[0059] 4. Crankshaft rod. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0061] Example 1

[0062] The technical solution in this application embodiment is to effectively solve the technical problem that existing crankshaft installation equipment cannot operate fully automatically during use. The overall idea is as follows:

[0063] refer to Figures 1 to 5 As shown, to address the problems existing in the prior art, the present invention provides a crankshaft clamping machine for achieving fully automatic crankshaft installation, thereby changing the traditional method that requires manual intervention. The clamping machine consists of four parts: a clamping mechanism 3 for clamping the crankshaft and a translation drive device for driving the clamping mechanism 3 to move horizontally, as detailed below:

[0064] The clamping mechanism 3 is mainly based on two corresponding clamping parts 32, namely pneumatic grippers 321. Since the connecting rod journals at both ends of the crankshaft need to be in contact with the edges of the cylinder block 2 during installation, directly gripping both ends of the crankshaft would easily cause the pneumatic grippers 321 to collide with the ends of the cylinder block 2. Therefore, it can only grip the connecting rod journals on the crankshaft. Furthermore, since the connecting rod journals on the crankshaft are arranged opposite each other, that is, multiple connecting rod journals are distributed on the upper and lower sides of the crankshaft, such as... Figure 5 As shown, therefore, for ease of gripping and to avoid collision with cylinder 2, the pneumatic gripper 321 grips the upward-facing connecting rod journal, as shown. Figure 5 As shown; based on this, we set two pneumatic grippers 321 on both sides of the docking seat 31, and made the distance between the two pneumatic grippers 321 consistent with the distance between the connecting rod journals at both ends of the crankshaft. In this way, when the pneumatic grippers 321 move towards the crankshaft, they can grip the connecting rod journals on the crankshaft accordingly.

[0065] like Figure 4As shown, the pneumatic gripper 321 mainly uses a common gripper cylinder available on the market. However, it is necessary to connect an extension claw 324 to each of its two movable claws 323. The extension claw 324 is restricted to the fixed clamp 322. The fixed clamp 322 and the movable claws 323 are connected by a rotating shaft. In this way, when the pneumatic gripper 321 drives the movable claws 323 to open and close, the power is transmitted from the movable claws 323 to the extension claws 324, thereby extending the gripping distance and length of the movable claws 323 to facilitate the gripping of the crankshaft. A flexible pad 325 made of rubber material is connected to one side of the two extension claws 324 to improve the fit between the pneumatic gripper 321 and the connecting rod journal.

[0066] The translation drive device mainly uses two parallel columns 11 as the support structure of the whole device, and connects a horizontally extending crossbeam 12 to it to provide a foundation for the lateral movement of the crankshaft. Then, a servo slide with the same length direction is installed on the crossbeam 12, and the pneumatic gripper 321 is installed on the servo slide. Therefore, when the servo slide is working, it can drive the pneumatic gripper 321 to move horizontally. In addition, in order to grasp, a lifting rod 13 that can move up and down is installed on the servo slide, and the pneumatic gripper 321 is installed at the bottom of the lifting rod 13. In this way, the pneumatic gripper 321 can have the function of lifting under the action of the lifting rod 13, thereby performing grasping and placing actions. Since we need horizontal movement and lifting functions, it is not limited to the above-mentioned structure. Other devices with the same function or structure can be used, such as the moving structure given in patent: CN121083605A, etc.

[0067] Furthermore, since the direction of movement of the cylinder block 2 is perpendicular to the extension direction of the crossbeam 12 during crankshaft installation (i.e., longitudinal), while the crankshaft is moved laterally during loading and unloading, it is necessary to rotate the crankshaft. To this end, we connect a rotary cylinder 14 to the bottom of the lifting rod 13 so that it can dock with the docking seat 31. In this way, the rotary cylinder 14 can drive the crankshaft rod 4 to rotate 90°, thereby completing the rotation of the crankshaft.

[0068] In the specific implementation process, it is divided into two steps. The first step is gripping. The servo slide installed on the crossbeam 12 drives the lifting rod 13 and the pneumatic gripper 321 on the lifting rod 13 to move laterally until the pneumatic gripper 321 is above the crankshaft to be installed. Then, the lifting rod 13 is controlled to descend, and at the same time, the extension claw 324 at the bottom of the pneumatic gripper 321 unfolds. When the connecting rod journal on the crankshaft enters between the two extension claws 324, the downward movement stops. At the same time, the pneumatic gripper 321 is controlled to drive the extension claw 324 to close, thereby clamping the connecting rod journal of the crankshaft. Finally, the lifting rod 13 is controlled to move the crankshaft upward, thus completing the gripping of the crankshaft.

[0069] The second step is assembly. The servo slide on the crossbeam 12 is controlled again to move the crankshaft to the other end of the crossbeam 12 until the crankshaft is directly above the cylinder block 2. At this time, the rotary cylinder 14 is controlled to drive the pneumatic gripper 321 and the crankshaft being gripped to rotate 90°, so that the crankshaft is aligned with the crankshaft position on the cylinder block 2. Then, the lifting rod 13 is controlled to move the crankshaft down and place it on the cylinder block 2. Finally, the pneumatic gripper 321 is controlled to open and move up through the lifting rod 13, thus completing the crankshaft assembly.

[0070] Example 2

[0071] Based on Example 1, this application provides a control method for a crankshaft clamping machine, the overall concept of which is as follows:

[0072] After the equipment is powered on and the air supply is connected, the operator presses the reset button on the control panel. The logic controller simultaneously sends reset commands to the horizontal servo slide, lifting rod, rotary cylinder, and pneumatic gripper, controlling each component to return to its preset initial position. Specifically, the horizontal servo slide returns to the initial coordinate X0 directly above the crankshaft loading station, the lifting rod returns to the highest safe height Z0, the rotary cylinder returns to the initial angle of 0°, and the pneumatic gripper returns to its fully open initial state. Simultaneously, a full-dimensional self-check is performed, checking the equipment's air supply pressure, the operating status of each actuator, the end-to-end communication status, the communication status of the upstream and downstream control systems, and the operating status of the edge computing nodes. When all test items meet the preset standards and all actuators have received their return-to-position signals, the equipment enters the standby state, and the automatic operation indicator lights up. If any test item fails, the corresponding audible and visual alarm is immediately triggered, the fault code and troubleshooting instructions are displayed on the control panel, and the automatic operation permission of the equipment is locked, allowing only manual mode for troubleshooting.

[0073] Once the equipment enters the standby state, the logic controller acquires comprehensive data related to the single crankshaft to be assembled, the matching engine block, and the corresponding complete engine through upstream and downstream supporting management and control systems. This includes data on crankshaft machining dimensions, form and position tolerances, dynamic balance testing, and material performance output from the upstream crankshaft machining process management system; matching cylinder block machining dimensions, assembly tolerances, and bearing performance parameters output from the enterprise's product lifecycle management system; complete engine design performance indicators and subsequent assembly requirements output from the downstream engine assembly management and control system; and typical service condition load data for the corresponding engine model output from the vehicle service condition management and control platform. This comprehensive data is then transmitted in real time to the edge computing node, which... A dedicated assembly adaptation control model for this single crankshaft was constructed. Based on the finite element analysis method, the model simulates the stress distribution, deformation, and assembly coaxiality of the crankshaft under different clamping positions, clamping forces, rotation angles, and descent speeds. Combining the cylinder block assembly datum, subsequent process requirements, and vehicle service conditions, the optimal assembly control parameters for this crankshaft were calculated to avoid the risks of assembly stress concentration, tolerance accumulation in subsequent processes, and abnormal wear during service. Finally, a personalized control parameter set uniquely bound to this single crankshaft was generated. The parameter set specifically includes the target clamping journal position, target clamping force control curve, gripping alignment coordinates, assembly alignment coordinates, target rotation angle control curve, graded lifting stroke, graded running speed, and clamping and holding pressure duration.

[0074] While constructing the assembly adaptation control model, the logic controller synchronously acquires data on the current ambient temperature, humidity, air source temperature, air source dew point, power grid voltage fluctuations, and cumulative equipment runtime through the workshop sensing system. This data is then input into the assembly adaptation control model to perform multi-dimensional dynamic correction and compensation on the personalized control parameter set. Specifically, based on ambient temperature and humidity data, the output pressure compensation coefficient of pneumatic components is corrected to offset the impact of temperature and humidity on pneumatic component output; based on cumulative equipment runtime data, the positioning coordinates of the servo system are corrected to offset the accuracy degradation caused by long-term equipment wear; and based on power grid voltage fluctuation data, the speed loop parameters of the servo system are corrected to avoid operating speed deviations caused by voltage fluctuations. Ultimately, this offsets the impact of environmental variables and equipment wear on assembly accuracy.

[0075] For different application scenarios, the control parameters were simultaneously adapted and adjusted: For the S04 passenger car general-purpose machine, the positioning accuracy threshold of the horizontal servo slide and lifting rod was set to ±0.03mm, and the single-cycle operation target was set to ≤50s to meet the production needs of large-volume, high-cycle operation; For the S6000 heavy-duty commercial vehicle machine, the clamping force control accuracy threshold of the pneumatic gripper was set to ±1%, and the lifting speed was reduced to 60% of the conventional value to meet the needs of stable clamping and assembly of heavy crankshafts; For the new energy hybrid dedicated high-efficiency machine, the crankshaft attitude matching accuracy threshold was set to ±0.02°, and a superposition verification step of crankshaft residual stress and assembly stress was added to meet the performance requirements of high speed and high reliability.

[0076] After receiving the crankshaft loading trigger signal from the crankshaft conveyor, the logic controller retrieves the corresponding personalized control parameter group for that crankshaft. Based on the gripping alignment coordinates, it controls the horizontal servo slide to move the lifting rod horizontally to directly above the target crankshaft journal. Then, based on the graded lifting stroke, it controls the lifting rod to descend to the preset gripping height, ensuring the pneumatic gripper's gripping extension arms completely enclose the target journal. Subsequently, it controls the pneumatic gripper's gripping extension arms to close in opposite directions, dynamically adjusting the output pressure of the pneumatic gripper according to the target gripping force control curve, executing the gripping force closing... The system employs a PID control loop. Initially, the gripper closes rapidly at 30% of the target pressure. After contacting the journal, the pressure gradually increases to the target gripping force, eventually stabilizing within ±2% of the target gripping force. Only after receiving a proximity switch feedback signal indicating that the gripper is in place, and upon receiving both signals that the real-time gripping force deviates from the preset value within ±2% of the threshold, the system controls the lifting rod to raise the crankshaft to the preset safe height, completing the gripping action. If either signal is not met, the system immediately controls the pneumatic gripper to open, re-executes the alignment and gripping action, and triggers an alarm if the gripping fails three times consecutively.

[0077] Throughout the entire operation, the logic controller obtains the action sequence plan and real-time air consumption data of all pneumatic equipment in the entire production line through the centralized management and control platform of the workshop pneumatic system. Based on the multi-agent reinforcement learning model, it predicts the air source pressure fluctuation trend and peak air consumption range in the next 30 seconds. Under the premise of not exceeding the 80-second cycle requirement of the production line, it actively adjusts the execution sequence window of the pneumatic action of the equipment to avoid the peak air consumption range in the entire area. At the same time, it dynamically adjusts the output pressure compensation coefficient of the pneumatic gripper and rotary cylinder to eliminate the impact of air source pressure fluctuation on clamping accuracy and rotary positioning accuracy.

[0078] After the grasping action is completed, the logic controller controls the horizontal servo slide to move the crankshaft horizontally to the assembly station according to the assembly alignment coordinates in the personalized control parameter group. The movement process adopts an S-shaped acceleration and deceleration curve to avoid crankshaft wobble caused by start-stop impact. Throughout the horizontal movement, the rotary cylinder is synchronously controlled to dynamically adjust the spatial attitude of the crankshaft according to the target rotation angle control curve, and corrects the crankshaft wobble and torsional errors in real time, realizing the parallel execution of transfer and attitude adjustment. The rotation angle is fed back in real time through the rotary cylinder encoder, and compared with the target rotation angle control curve in real time and dynamically corrects the output parameters so that when the crankshaft reaches the top of the assembly station, the spatial attitude is perfectly matched with the installation reference of the engine block, and the matching degree error is controlled within ±0.02°. The attitude adjustment action stops only after receiving a feedback signal that the attitude matching is qualified.

[0079] After receiving the cylinder block arrival trigger signal from the cylinder block conveying device, the logic controller first performs a pre-simulation of the entire crankshaft descent assembly process using the assembly adaptation control model. This simulates the relative positions of the crankshaft with the cylinder block and surrounding tooling during descent, identifying assembly interference risks. Based on the pre-simulation results, the controller corrects the staged lifting stroke and operating speed curves. Then, according to the corrected parameters, it controls the lifting rod to drive the crankshaft to descend in stages. During the descent, a laser rangefinder sensor collects real-time data on the distance between the lower end face of the crankshaft and the cylinder block mounting surface. When the distance exceeds the 50mm safety threshold, the lifting rod moves at a speed of 200m. The crankshaft descends rapidly at a first speed of m / s. When the distance is less than or equal to the safety threshold of 50mm, it automatically switches to a second speed of 20mm / s for a slow descent until the crankshaft is stably placed in the designated assembly position on the engine block. Only after receiving the pressure feedback signal indicating that the crankshaft is in place and the dual signals indicating that the lifting rod has reached the target lowering coordinate, does the pneumatic gripper fully open to release the crankshaft. Then, the lifting rod is controlled to rise back to the preset safe height to complete the assembly action. If the placement signal is not triggered, the descent action is immediately stopped to check for abnormalities and avoid a hard collision between the crankshaft and the engine block.

[0080] After the assembly operation is completed, the logic controller controls the rotary cylinder to rotate in the opposite direction to return to the initial position of 0°. Simultaneously, it controls the horizontal servo slide to drive the lifting rod back to the initial position directly above the crankshaft loading station, completing the single-cycle reset. At the same time, the actual operation data of the entire operation and the assembly status detection data are input into the assembly adaptation control model to complete the virtual-real calibration and iterative optimization of the model and correct the simulation deviation of the model. The calibrated assembly adaptation control model, personalized control parameter group and the entire operation data are encrypted with AES-256 and bound to the unique identification number of the corresponding crankshaft and cylinder block. They are then uploaded to the upstream and downstream supporting management and control system for closed-loop archiving. At the same time, the assembly adaptation control model construction rules of the same model crankshaft are optimized based on the data of this operation and stored in the dedicated database of the edge computing node to provide a preliminary optimization basis for the assembly control of the same model crankshaft.

[0081] While archiving the work data, the logic controller uploads the encrypted full-process work data, the calibrated assembly adaptation control model, and the operator's operation record data to the enterprise alliance blockchain platform for tamper-proof notarization based on privacy computing technology. This generates a notarization hash value that is strongly bound to the unique identification number of the corresponding crankshaft, cylinder block, and complete machine. Upstream processing enterprises, engine manufacturers, downstream vehicle manufacturers, and regulatory agencies can all use the unique identification number and notarization hash value to trace back the complete crankshaft clamping process data under the premise that the data is available but not visible, thus meeting the compliance management requirements of core components in the automotive industry.

[0082] Throughout the entire operation process, the logic controller performs dual-channel distributed mirror storage of the equipment's full-state operation data at a fixed sampling frequency of 100Hz. This means that the data is synchronously backed up in real time in both the logic controller's local storage unit and the edge computing node's storage unit. The full-state operation data includes real-time position data of each component, operating status data, clamping status data, crankshaft unique identification number, executed step data, unexecuted step data, and real-time communication data. When the equipment experiences an abnormal power outage or communication interruption, the logic controller completes the encrypted storage of the last full-state operation data through the UPS backup power supply. After the equipment returns to normal, it automatically verifies the integrity of the dual-channel stored data. Once it confirms that no data has been lost, it resumes the remaining operation steps from the execution breakpoint where the non-steady-state condition occurred, without needing to perform a full-process reset and rework. Meanwhile, the logic controller collects real-time data on the operating current, voltage, and vibration of the horizontal servo slide servo motor, the clamping force fluctuation data of the pneumatic gripper, and the action timing data of the rotary cylinder. It extracts the feature values ​​of the above multi-source heterogeneous data, inputs them into the pre-trained multimodal small-sample fault prediction model, and combines the equipment health digital twin to simulate the fault evolution trend to predict the early potential fault risks of the corresponding components. When a potential fault risk is predicted, the operating parameters of the corresponding components are automatically adjusted to reduce the load, and fault warning information and precise maintenance guidance are generated and uploaded to the enterprise's equipment lifecycle management system.

[0083] The logic controller establishes bidirectional data communication with the crankshaft conveyor, cylinder block conveyor, and upstream and downstream production equipment within the production line through a time-sensitive network. Based on the operating data with microsecond-level timestamps uploaded by each device, it executes microsecond-level timing synchronization control across the entire production line, ensuring precise matching between the crankshaft clamping process and the actions of upstream and downstream equipment. This completely eliminates the need for traditional hardware I / O trigger signal transmission, removing latency errors and electromagnetic interference caused by hardware triggers. Simultaneously, it establishes distributed cluster communication with three identical crankshaft clamping machines in the workshop through edge computing nodes. This allows for real-time acquisition of the cumulative runtime, real-time load rate, equipment health status, fault warning status, and number of pending tasks for each device within the cluster. Through distributed reinforcement learning algorithms, it autonomously performs global capacity optimization and intelligent load balancing scheduling for the entire production line, dynamically allocating pending crankshaft clamping tasks to maintain the load rate and cumulative runtime of each device within the cluster within a balanced range of ±5%. When a single device in the cluster triggers a fault warning, the remaining devices automatically take over the pending tasks, ensuring uninterrupted continuous production.

[0084] In addition, the logic controller continuously acquires subsequent full-process assembly and testing data, engine test performance data, real-time service condition data after vehicle assembly, and fault maintenance data of the engine corresponding to the crankshaft through the downstream engine assembly and control system and the vehicle service condition control platform. Based on the above full life cycle data, it constructs an association mapping model of assembly parameters-engine performance-service life, simulates the impact of different assembly parameters on engine service performance through digital twins, and iteratively optimizes the assembly adaptation control model construction logic and personalized control parameter generation rules of crankshafts of the same batch and model in reverse. The results are stored in a dedicated database to achieve continuous iterative optimization of assembly accuracy and engine service performance.

[0085] Example 3

[0086] This embodiment addresses the problems of pure data-driven black-box models, such as uninterpretability, poor generalization, output results that do not conform to the physical laws of assembly, and high false judgment rate with fixed thresholds. It refines the core model of this invention as follows:

[0087] First, a dedicated assembly adaptation control model for a single crankshaft was constructed. Initially, when building the model using the industry-standard LSTM neural network, only the crankshaft size data was input. The model achieved high prediction accuracy on the S04 general-purpose model, but the accuracy dropped significantly on the S6000 heavy-duty model and hybrid special-purpose model. Furthermore, abnormal outputs that did not conform to physical laws, such as assembly stress exceeding the allowable stress of the material, frequently occurred. Therefore, it was finally determined to build the model based on the Physical Information Neural Network (PINN), embedding physical constraints into the neural network to achieve dual-constraint prediction based on data-driven and physical laws.

[0088] Before model training, production and testing data from multiple engine manufacturers over the past five years were collected. After data cleaning, deduplication, and outlier removal, 15,000 valid samples were selected, covering 42 crankshaft models, 28 cylinder block models, and 32 environmental service conditions. Each sample set includes multi-dimensional operational data throughout the assembly cycle, actual performance degradation data within the corresponding three-year natural service cycle of the engine, and root cause annotation data for assembly failures and process defects. The samples were randomly divided into training, validation, and test sets in a 7:2:1 ratio to ensure that the engine models, cylinder blocks, and environments of the training, validation, and test sets are consistent, thus constructing a performance evolution training dataset.

[0089] The model's overall architecture consists of three core functional units. The first unit is a two-way coupling unit for crankshaft characteristics and assembly process, containing three sub-units: The substrate intrinsic property mapping sub-unit, based on a generalized Kelvin creep constitutive model, obtains creep constitutive equation parameters for crankshafts with different materials, structures, and residual stresses through 800 sets of uniaxial creep tests, constructs a crankshaft characteristic database, and builds a mapping model between crankshaft intrinsic parameters and creep constitutive equation parameters based on a gradient boosting tree algorithm. This model converts the input crankshaft machining and material performance data into input parameters for the creep constitutive equation of the assembly process, establishing a quantitative correlation between crankshaft intrinsic properties and deformation behavior during assembly. The assembly motion dynamics sub-unit establishes a rigid body dynamics model of the servo system-crankshaft for the entire process of crankshaft clamping, transporting, and placing. The formula is... Where q is the system's generalized coordinates, and M(q) is the mass matrix. Here, G(q) represents the Coriolis force and centrifugal force matrix, G(q) represents the gravity term, and τ represents the servo system driving torque. Model parameters for different crankshaft models are obtained through dynamic identification experiments, and a mapping database between crankshaft models and dynamic parameters is constructed. By inputting crankshaft parameters and assembly action commands, the dynamic response and deformation of the assembly process can be output in real time. The bidirectional coupling solution subunit incorporates a bidirectional differential equation for crankshaft assembly stress and action parameters. The output of the crankshaft creep constitutive equation and the output of the assembly dynamic model are iteratively solved. The iteration step size is set to 10ms, and the iteration convergence threshold is 1e-6. In each iteration, the assembly stress and deformation at the current moment are first calculated using the assembly dynamic model, and then the creep strain of the crankshaft is calculated using the creep constitutive equation. The strain result is then fed back to the dynamic model to correct the action parameters for the next iteration. Ultimately, this achieves bidirectional coupling calculation of crankshaft characteristics and the assembly process, outputting coupled characteristic evolution data.

[0090] The second unit is the assembly-service cross-cycle time-dependent unit, which contains three sub-units: The assembly stage division sub-unit divides the assembly process into three consecutive stages—alignment gripping, transport attitude adjustment, and graded descent placement—based on the changes in assembly stress, crankshaft deformation, and clamping force in the coupling feature evolution data. It acquires the state data output by the bidirectional coupling unit in real time, determines the current assembly stage based on the state interval, and assigns an independent feature extraction channel to each stage. The time-series feature extraction sub-unit builds a feature extraction network based on a gated recurrent unit (GRU), with a hidden layer dimension of 128 and two layers. It controls the transmission and forgetting of time-series information through update and reset gates. The position, force, and velocity time-series data of each assembly stage are input into the network to filter out time-series features valuable for subsequent performance prediction and output a high-dimensional feature vector for each assembly stage. The cross-cycle correlation mapping sub-unit is built on a bidirectional long short-term memory network (Bi-LSTM) to construct a time-series correlation network with 3 layers and 256 hidden layer dimensions. It can learn the dependencies of time-series sequences from both forward and backward directions. The feature vectors of the three assembly stages are concatenated into a complete time-series feature sequence and input into the network to establish a nonlinear mapping relationship between the key feature time-series sequences of the assembly stages and the performance degradation indicators of the service cycle, and output the performance degradation prediction results of the crankshaft within 3 years of its service cycle.

[0091] The third unit is the assembly interference and failure risk identification unit, which contains three sub-units: The parameter disturbance analysis sub-unit, with a step size of 10ms, traverses all time nodes of the assembly process, applying two sets of disturbances (+5% and -5%) to the key action parameters of each node, respectively. These disturbances are input into the bidirectional coupling unit and the cross-cycle time-dependent unit to obtain the coupling characteristic evolution data and performance degradation prediction results after the disturbance, saving two sets of comparative data before and after the disturbance for each time node; The failure amplification effect calculation sub-unit calculates the corresponding failure amplification coefficient based on the data before and after the disturbance at each time node, using the formula: Failure amplification coefficient = |probability of failure after disturbance - The failure probability before disturbance is calculated as | / failure probability before disturbance × 100%, generating a time-series curve of the failure amplification factor for the entire assembly cycle. The risk interval positioning subunit determines the time interval with a failure amplification factor ≥ 20% as the assembly failure sensitive feature interval, locates the chaotic sensitive boundary region of the assembly stage corresponding to this interval, and classifies the sensitivity level according to the average failure amplification factor within the interval: above 100% is an extremely high sensitivity region, 50%-100% is a high sensitivity region, and 20%-50% is a medium sensitivity region. Different control precision and acquisition density are matched for different sensitivity levels, and the start time, end time, range, and sensitivity level of the failure sensitive feature interval are output.

[0092] The model was trained under supervised training using the Adam optimizer, employing a dual-constraint loss function. The total loss function formula was Loss = Loss_data + λ × Loss_physics, where Loss_data is the mean squared error between the measured performance degradation data and the model's prediction results, and Loss_physics is the computational residual of the bidirectional coupled physical equations. The regularization coefficient λ was optimized within the range of 0.05-0.2 using cross-validation, and was ultimately set to λ = 0.1. The training hyperparameters were set to 600 iterations, an initial learning rate of 0.001, a 10% decay every 100 iterations, and a batch size of 32. An early stopping strategy was adopted, terminating training early when the validation set loss did not decrease for 50 consecutive iterations. The model was considered converged when the model's loss function on the validation set converged to ≤0.01 and the performance degradation prediction accuracy on the test set was ≥96%. The model was then deployed to an industrial control computer in ONNX format.

[0093] A counterfactual reasoning-based fault prediction and root cause tracing model was simultaneously built. Based on a structural causal model framework, historical normal operation samples, fault samples, and corresponding root cause annotation data were first collected to construct a basic sample library, with a normal sample to fault sample ratio of 7:3. The model includes a counterfactual reasoning unit, a fault root cause localization unit, a long-term failure prediction unit, and an incremental adaptation unit. The counterfactual reasoning unit constructs a counterfactual reasoning framework based on the Do operator in causal science, fixing all other features and applying intervention operations only to a single target causal feature. This constructs a counterfactual scenario of feature intervention followed by fault occurrence, calculating the change in the probability of fault occurrence before and after the feature intervention as the feature's contribution to the equipment failure. After traversing all causal features, the contribution ranking of all features is obtained. The fault root cause localization unit sorts each feature from highest to lowest contribution, placing features with a contribution ≥ 40% of the features are identified as the core root causes of equipment failure, while features contributing 20%–40% are identified as secondary influencing factors. The output includes root cause localization results, including root cause type, feature deviation value, influence weight, and maintenance optimization direction. The long-term failure prediction unit combines equipment operation trends and feature deviation values, and uses a pre-trained Bayesian probability model to calculate the probability of equipment failure at different future time points, outputting the comprehensive failure probability and failure mode probability distribution over a 3-month period. The incremental adaptation unit is built based on a small-sample learning algorithm of a prototype network. Through the causal feature extraction capability learned by the pre-trained model, it constructs feature prototypes for small samples of new scenarios. Only ≥50 sets of small-sample data for new scenarios are needed to complete the rapid adaptation of the model without retraining the entire dataset. After the model pre-training is completed, the model converges when the root cause localization accuracy on the test set is ≥92%.

[0094] A multi-factor causal-driven dynamic dual-threshold decision model was simultaneously built, based on a causal-enhanced gradient boosting tree algorithm. The model incorporates a causal constraint module and a causal-enhanced gradient boosting tree prediction unit. The causal constraint module, based on a directed acyclic graph (DAG) causal graph algorithm, constructs a causal graph between scenario parameters, service failure probability, and decision thresholds. Independent condition checks are used to identify core factors with direct causal relationships to threshold adjustments, while eliminating interfering factors that only exhibit correlation without causal relationships. The causal-enhanced gradient boosting tree prediction unit adds a causal constraint regularization term to the loss function of a traditional gradient boosting tree, ensuring that the model's split nodes prioritize core causal factors. The model inputs include environmental operating parameters, crankshaft intrinsic parameters, overall machine performance requirements, and service cycle failure probability. The outputs are the upper limit of the pass / fail decision threshold and the lower limit of the fail / fail decision threshold. Pass / fail and fail decision thresholds validated through long-term service in historical scenarios were collected to construct a threshold decision training dataset. This dataset covers relevant data from 42 crankshaft models, 28 cylinder block models, and 32 environmental service conditions. Five-fold cross-validation was used to train and optimize the model. Model convergence was achieved when the threshold prediction deviation was ≤5%.

[0095] Finally, it should be noted that the above embodiments are merely examples for clearly illustrating the present invention and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A control method for a crankshaft clamping machine, characterized in that, The crankshaft clamping machine includes a truss and two sets of pneumatic grippers for gripping the crankshaft. A crossbeam is mounted on the top of the truss, and the crossbeam has a horizontal servo slide that can reciprocate horizontally along its length. A lifting rod is mounted on the horizontal servo slide, which can move horizontally synchronously with the horizontal servo slide and reciprocate vertically. The two sets of pneumatic grippers are respectively located on both sides of the bottom of the lifting rod and connected to a rotary cylinder at the bottom of the lifting rod. The crankshaft clamping machine also includes a logic controller, which is electrically connected to the horizontal servo slide, the lifting rod, the rotary cylinder, and the pneumatic grippers, and is used to send control commands and receive operating status feedback signals. The logic controller establishes bidirectional encrypted data communication with a supporting edge computing node and a supporting upstream and downstream management and control system of the industrial chain. This control method includes the following steps: S1. Return each execution component to its preset initial position. After the device performs a full-dimensional self-check and all components are confirmed to be in place, it enters the standby state. S2. Obtain the full-dimensional correlation data of the single crankshaft to be assembled, the matching engine block, and the corresponding whole machine. After transmitting the data to the edge computing node for processing, generate a personalized control parameter group that is uniquely bound to the single crankshaft. S3. After receiving the trigger signal that the crankshaft is in place, control the horizontal servo slide and lifting rod according to the personalized control parameter group to complete the gripping and alignment, and then control the pneumatic gripper to complete the adaptive gripping of the crankshaft. After the gripping state meets the preset requirements, control the lifting rod to drive the crankshaft to rise to a safe height and complete the gripping action. S4. According to the personalized control parameter group, control the horizontal servo slide to drive the clamped crankshaft to the assembly station for transfer. During the transfer, control the rotary cylinder to dynamically adjust the spatial attitude of the crankshaft and correct the attitude deviation so that when the crankshaft reaches the top of the assembly station, the attitude matches the cylinder mounting reference. S5. After receiving the trigger signal that the cylinder block is in place, control the lifting rod to drive the crankshaft to descend in stages according to the personalized control parameter group, so as to complete the stable placement of the crankshaft in the designated position of the engine cylinder block. After the placement is in place, control the pneumatic gripper to release the crankshaft, and the lifting rod to rise to a safe height to complete the assembly action. S6. Control all execution components to return to their initial positions, complete the encrypted archiving of the entire process data of this operation, and simultaneously perform iterative optimization of the control logic based on the actual operation data of this operation, waiting for the next operation trigger signal.

2. The control method of the crankshaft clamping machine according to claim 1, characterized by, In step S1, the device's full-dimensional self-test specifically includes: The logic controller performs item-by-item detection on the equipment's air supply pressure, the operating status of each actuator, the communication link status of each component, the communication status of the upstream and downstream supporting control systems in the industrial chain, and the operating status of the edge computing nodes. Once all test results meet the preset standards and a feedback signal indicating that all actuators have returned to their initial positions is received, the equipment enters the standby state. If the self-test fails, a fault alarm for the corresponding item will be triggered and automatic operation privileges will be locked.

3. The control method of the crankshaft clamping machine according to claim 1, characterized by, In step S2, the full-dimensional correlation data of the single crankshaft to be assembled, the matching engine block, and the corresponding whole machine includes the full-process machining dimension data, form and position tolerance data, dynamic balance test data, and material performance test data of the single crankshaft; the machining dimension data, assembly tolerance data, and bearing performance parameters of the matching engine block; and the design performance indicators and subsequent assembly requirements of the corresponding whole machine. The edge computing node constructs a dedicated assembly adaptation control model for the single crankshaft using the aforementioned multi-dimensional associated data. Through simulation calculations of the assembly adaptation control model, it generates the personalized control parameter set, which includes the target clamping journal position, target clamping force control curve, gripping alignment coordinates, assembly alignment coordinates, target rotation angle control curve, graded lifting stroke, graded running speed curve, and clamping and holding pressure duration.

4. The control method of the crankshaft clamping machine according to claim 3, characterized in that, In step S2, when the logic controller constructs the assembly adaptation control model for a single crankshaft, it simultaneously obtains data on the current ambient temperature, ambient humidity, air source temperature, air source dew point, grid voltage fluctuation, and cumulative equipment operating time through the workshop sensing system. The above data is then input into the assembly adaptation control model to perform multi-dimensional dynamic correction and compensation on the personalized control parameter group, thereby offsetting the impact of environmental variables and equipment wear on assembly accuracy. Throughout the entire operation, the logic controller acquires the action sequence plan and real-time air consumption data of all pneumatic equipment in the production line, predicts the air source pressure fluctuation trend and peak air consumption range, and adjusts the execution sequence window of the pneumatic action of the equipment under the premise of meeting the production line cycle requirements, while dynamically adjusting the output pressure compensation coefficient of the pneumatic components.

5. The control method of the crankshaft clamping machine according to claim 1, characterized by, The specific steps of step S3 are as follows: After the logic controller receives the crankshaft loading trigger signal sent by the crankshaft conveying device, it controls the horizontal servo slide to move the lifting rod horizontally to the gripping position directly above the crankshaft target clamping journal according to the personalized control parameter group, and then controls the lifting rod to descend to the preset gripping height according to the graded lifting stroke. Then, the clamping extension arms of the pneumatic gripper are controlled to close in opposite directions, and the clamping force is dynamically adjusted according to the target clamping force control curve in the personalized control parameter group to clamp the target clamping journal position of the crankshaft. Upon receiving dual feedback signals indicating that the clamping position is in place and the clamping force deviates from the preset value within the threshold range, the control lifting rod drives the clamped crankshaft to rise to the preset safe height, completing the gripping action.

6. The control method for the crankshaft clamping machine according to claim 1, characterized in that, Specifically, step S4 involves the logic controller controlling the horizontal servo slide to move the clamped crankshaft horizontally to the assembly station according to the personalized control parameter group. Throughout the horizontal movement, the synchronously controlled rotary cylinder adjusts the spatial attitude of the crankshaft according to the target rotation angle control curve within the personalized control parameter group, and corrects the crankshaft's runout and torsional errors in real time. The attitude adjustment action stops only after the crankshaft reaches directly above the assembly station and a feedback signal is received indicating that the crankshaft attitude matches the preset requirements for the cylinder block mounting reference.

7. The control method of the crankshaft clamping machine according to claim 1, characterized by, The specific steps of step S5 are as follows: After the logic controller receives the cylinder block arrival trigger signal sent by the cylinder block conveying device, it performs a full-process pre-simulation of the crankshaft lowering assembly process through the assembly adaptation control model, checks for assembly interference risks, and corrects the staged lifting stroke and running speed curves. According to the revised personalized control parameter group, the control lift rod drives the crankshaft to descend in stages. During the descent, the distance data between the lower end face of the crankshaft and the cylinder block mounting surface is collected in real time. When the distance is greater than the preset safety threshold, the crankshaft descends at the first speed. When the distance is less than or equal to the preset safety threshold, the crankshaft descends at the second speed. The second speed is less than the first speed until the crankshaft is stably placed in the designated assembly position of the cylinder block. Only after receiving a feedback signal that the crankshaft has been placed in place will the pneumatic gripper's clamping extension arm be opened to release the crankshaft, and then the lifting rod be raised back to the preset safe height to complete the assembly action.

8. The control method of the crankshaft clamping machine according to claim 1, characterized by, Specifically, step S6 involves the logic controller controlling the rotary cylinder, horizontal servo slide, and lifting rod to return to their initial positions in sequence; simultaneously inputting the actual operation data and assembly status detection data of the entire operation into the assembly adaptation control model for virtual-real calibration and iterative optimization of the model. After encrypting the calibrated assembly adaptation control model, personalized control parameter group, and full-process operation data, they are bound to the unique identification number of the corresponding crankshaft and cylinder block, and simultaneously uploaded to the upstream and downstream supporting management and control system of the industry chain for archiving. At the same time, based on the data from this operation, the assembly adaptation control model construction rules for crankshafts of the same model are optimized, and the operation trigger signal for the next crankshaft is awaited.

9. The control method of the crankshaft clamping machine according to claim 8, characterized in that, The logic controller establishes bidirectional data communication with the crankshaft conveying device, cylinder block conveying device, and upstream and downstream production equipment in the production line through a time-sensitive network. All connected devices use a unified clock source to generate microsecond-level timestamps. Based on the timestamped operating data, the controller performs microsecond-level timing synchronization control across the entire production line, ensuring that the crankshaft clamping process matches the actions of upstream and downstream equipment on the production line. Meanwhile, the logic controller establishes distributed cluster communication with all crankshaft clamping machines of the same model in the workshop through edge computing nodes, and obtains the cumulative runtime, real-time load rate, equipment health status and number of tasks to be executed for each device in the cluster in real time. It dynamically allocates clamping tasks to be executed through a collaborative scheduling algorithm, keeping the load rate and cumulative runtime of each device in the cluster within a preset balance range. When archiving a job, the logic controller uploads the encrypted full-process job data to the consortium blockchain platform for tamper-proof notarization, generating a notarization hash value that is strongly bound to the unique identification number of the corresponding crankshaft, cylinder block, and whole machine.

10. The control method of the crankshaft jacking machine according to claim 1, characterized by, Throughout the entire operation process from steps S1 to S6, the logic controller performs dual-channel distributed mirror storage of the equipment's full-state operation data at a preset fixed sampling frequency. The full-state operation data includes real-time position data of each component, operation status data, clamping status data, crankshaft unique identification number, executed step data, unexecuted step data, and real-time communication data. When the equipment experiences an abnormal power outage or communication interruption, the logic controller uses the backup power supply to encrypt and store the current status data. After the equipment returns to normal, it automatically verifies the integrity of the dual-channel stored data. Once confirmed, it continues to complete the remaining operation steps from the execution breakpoint where the abnormal operation occurred. Meanwhile, the logic controller collects the operating electrical signals, vibration signals, action timing signals, and force feedback signals of each actuator in real time, extracts signal feature values ​​and inputs them into the pre-trained fault prediction model to identify early potential fault risks of the components, adjusts the operating parameters of the corresponding components to reduce the load, and triggers the generation of fault warnings and maintenance guidelines and uploads them to the enterprise equipment management system.

Images