A high-precision position adjustment method in radiator core assembly

Abstract

This invention relates to the field of high-precision motion control technology for radiator assembly, and discloses a high-precision position adjustment method in radiator core assembly, comprising: acquiring the real-time spatial position feedback sequence of the displacement adjustment mechanism and calculating the synchronization deviation; establishing a reduced-order expansion state observer to extract the disturbance estimate of the assembly contact resistance; determining the real-time virtual admittance coefficient in combination with preset core stiffness parameters; determining the micro-position yield compensation amount and correcting the sliding mode convergence target based on the disturbance estimate and the real-time virtual admittance coefficient when the preset deviation threshold is met. This invention establishes a dynamic matching rule between the position adjustment amount and the internal stress evolution of the core, enabling the displacement adjustment mechanism to generate a controlled and compliant yield displacement during the assembly and fitting stage, actively dissipating the accumulated elastic potential energy, eliminating mechanical bounce and position overshoot caused by rigid waiting, and avoiding damage to the heat sink fins.

Description

Technical Field

[0001] This invention relates to a high-precision position adjustment method in the assembly of a radiator core, belonging to the field of high-precision motion control technology for radiator assembly. Background Technology

[0002] Currently, the heat exchanger manufacturing field adopts a multi-axis synchronous control strategy to complete the core pressing. The spatial trajectory coordination of multiple propulsion ends is achieved through displacement sensing sequence and servo drive loop. The control unit drives the power mechanism to generate displacement towards the assembly endpoint according to the preset position command. With the evolution of thermal management system towards high density and thin wall, the physical characteristics of heat dissipation fins exhibit nonlinear elastoplastic characteristics. During multi-axis synchronous extrusion, the yield strength of the fins fluctuates irregularly with the pressing depth. When faced with such nonlinear disturbances, the position adjustment system has difficulty decoupling the evolution relationship between position deviation and contact stress.

[0003] To maintain multi-end position synchronization, conventional solutions employ a synchronization tolerance dead zone. When any propulsion end experiences a displacement overshoot, the controller stops issuing position commands to that end. This method ignores the energy accumulation effect of the controlled object. During the synchronization waiting phase, the overshoot propulsion end continuously bears the reaction force generated by fin deformation. When the controller resumes issuing commands, the previously accumulated elastic potential energy is released instantaneously, causing displacement jumps and mechanical overshoot at the propulsion end. Furthermore, the software control logic has shortcomings. For example, Chinese invention patent application CN119781318A discloses an assembly control system for a radiator. The system, electronic equipment, and control methods obtain feedback and adjust the screw fastening gap through an external pressure detection component. Under the high-density pressing condition of the heat sink core, the nonlinear yielding characteristics of the fins exhibit high time-varying characteristics. The step adjustment logic based on the static pressure threshold is difficult to handle the submicron-level dynamic response requirements. Due to the dependence on external high-precision force sensing components, the system complexity and hardware cost are increased. The existing logic cannot perceive and predict the accumulation law of elastic potential energy inside the material. The transient of multi-axis synchronous switching cannot avoid the physical jump caused by stress release, making it difficult to achieve non-destructive precision assembly of ultra-thin-walled, high-density heat sink cores.

[0004] Therefore, how to determine a position adjustment mechanism that can decouple nonlinear structural resistance in real time and generate displacement yielding during the synchronous waiting phase to dissipate and accumulate energy, thereby avoiding position jumps at the moment of state switching, has become the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems in the background art, the technical solution of the present invention is as follows: A high-precision position adjustment method in the assembly of a radiator core, comprising the following steps: Step S1: Obtain the real-time spatial position feedback sequence of multiple radiator core displacement adjustment mechanisms, and calculate the independent position deviation of each radiator core displacement adjustment mechanism and the relative synchronization deviation between adjacent radiator core displacement adjustment mechanisms. Step S2: Establish a reduced-order extended state observer and extract the disturbance estimate reflecting the assembly contact resistance based on the real-time spatial position feedback sequence. The reduced-order extended state observer uses the system nominal model to obtain the acceleration term deviation caused by the external resistance. Step S3: Obtain the gradient of the change in the disturbance estimate, and determine the real-time virtual admittance coefficient based on the gradient and the preset core stiffness parameters. Step S4: When the independent position deviation of the specific radiator core displacement adjustment mechanism is reduced to within 0.05mm and the maximum global synchronization error exceeds 0.15mm, calculate the product of the disturbance estimate and the real-time virtual admittance coefficient, and determine the calculation result as the micro position yield compensation amount. Step S5: Adjust the convergence target position of the sliding mode approach law by using the micro position yield compensation amount, drive the heat sink core displacement adjustment mechanism to generate displacement that follows the fluctuation of assembly contact resistance, until the maximum value of global synchronization error decays to below 0.15mm, then adjust the real-time virtual admittance coefficient to zero value to restore tracking to the initially set target.

[0006] Preferably, step S1 includes the following steps: using a multi-branch state machine to determine the current assembly phase, which includes a synchronization waiting phase and a tracking phase; when it is determined that the current phase is a synchronization waiting phase, the product of the real-time virtual admittance coefficient and the relative synchronization deviation is determined as the micro position yield compensation amount, so that the heat sink core displacement adjustment mechanism generates a controlled elastic displacement.

[0007] Preferably, step S2 includes the following steps: performing a reduction-order operation on the acceleration state of the radiator core displacement adjustment mechanism, filtering out the second-order differential terms in the high-frequency noise components, and obtaining the disturbance estimate from the real-time spatial position feedback sequence as the input observation.

[0008] Preferably, step S5 includes the following steps: modulating the switching term gain in the sliding mode reaching law using the real-time virtual admittance coefficient, and increasing the system damping by increasing the real-time virtual admittance coefficient when the disturbance estimate exceeds the preset linear domain threshold, so as to suppress the transient chattering of the heat sink core displacement adjustment mechanism at the yield point of the heat sink fins.

[0009] Preferably, at the instant when the multi-branch state machine determines that the assembly phase has changed from the synchronous waiting phase to the tracking phase, the integral term coefficient in the control loop is adjusted according to the value of the disturbance estimate, and the historical accumulation path of the error is cut off by setting the integral term coefficient to zero.

[0010] Preferably, in step S2, the mechanical friction force is decoupled online using a reduced-order extended state observer, the contact force component reflecting the assembly interference intensity is separated from the disturbance estimate, and the contact force component is determined as the sole input source for the real-time virtual admittance coefficient calculation.

[0011] Preferably, the method utilizes real-time virtual admittance coefficients to establish dynamic matching rules between position adjustment amount and internal stress evolution of the core, generating controlled flexible displacement when the radiator core is closed.

[0012] Preferably, in step S3, the core stiffness parameters are preset according to the physical specifications of the heat sink to be assembled, and the response frequency matching the stiffness of the heat pipe array is obtained by adjusting the mapping slope between the real-time virtual admittance coefficient and the changing gradient.

[0013] Preferably, step S5 includes the following steps: sending the corrected convergence target position to the displacement adjustment driver, driving the displacement adjustment motor to drive the assembly displacement carrier to perform three-axis synchronous displacement compensation, and using the real-time virtual admittance coefficient to offset the relative synchronous deviation between multiple axes.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. In high-precision position adjustment, the nonlinear structural resistance and mechanical clearance disturbance during the assembly process are extracted and dynamically compensated in real time by the reduced-order extended state observer. This enables the position adjustment control quantity to synchronously offset the transient resistance fluctuations caused by the yielding of the heat sink fins. Under the condition that no external force sensing components are required, the position tracking trajectory at the end of the assembly execution is dynamically matched with the actual mechanical load characteristics in real time, avoiding workpiece stress damage caused by the mismatch between rigid control commands and nonlinear resistance.

[0015] 2. By utilizing the dynamic admittance evaluation function, the global synchronization deviation of multiple terminals is mapped in real time as a micro-position yield compensation value. This enables the advanced execution end to exhibit controlled physical elasticity characteristics when the synchronization waiting condition is met. By actively dissipating the elastic potential energy accumulated during the assembly process, the inherent method of single locking position command in traditional position adjustment is broken. This ensures that the execution end maintains the smoothness of the motion trajectory during the instantaneous transition from the waiting phase to the tracking phase, eliminating position overshoot and mechanical bounce phenomena caused by the instantaneous release of internal stress in the radiator core.

[0016] 3. The topology suppression link dynamically adjusts the integral gain coefficient of the term in the control loop based on the real-time judgment results of the multi-branch state machine. When multiple execution ends are in the synchronous misalignment interval, it effectively cuts off the historical accumulation path of the error. From the underlying logic of the control law, it avoids the risk of integral saturation of the actuator when it breaks through the resistance critical point. It also enhances the convergence stability of the system under complex time-varying load disturbances and ensures the continuity of assembly position adjustment commands in the spatial and temporal dimensions. Attached Figure Description

[0017] Figure 1 This is a flowchart of the heat sink core assembly position adjustment and controlled compliant yielding method of the present invention; Figure 2 This is a schematic diagram of the heat sink core assembly position adjustment system architecture and signal interaction principle of the present invention.

[0018] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0020] A high-precision position adjustment method in assembling a heat sink core includes the following steps: Step S1: Obtain the real-time spatial position feedback sequence of multiple radiator core displacement adjustment mechanisms, and calculate the independent position deviation of each radiator core displacement adjustment mechanism and the relative synchronization deviation between adjacent radiator core displacement adjustment mechanisms. Step S2: Establish a reduced-order extended state observer and extract the disturbance estimate reflecting the assembly contact resistance based on the real-time spatial position feedback sequence. The reduced-order extended state observer uses the system nominal model to obtain the acceleration term deviation caused by the external resistance. Step S3: Obtain the gradient of the change in the disturbance estimate, and determine the real-time virtual admittance coefficient based on the gradient and the preset core stiffness parameters. Step S4: When the independent position deviation of the specific radiator core displacement adjustment mechanism is reduced to within 0.05mm and the maximum global synchronization error exceeds 0.15mm, calculate the product of the disturbance estimate and the real-time virtual admittance coefficient, and determine the calculation result as the micro position yield compensation amount. Step S5: Adjust the convergence target position of the sliding mode approach law by using the micro position yield compensation amount, drive the heat sink core displacement adjustment mechanism to generate displacement that follows the fluctuation of assembly contact resistance, until the maximum value of global synchronization error decays to below 0.15mm, then adjust the real-time virtual admittance coefficient to zero value to restore tracking to the initially set target.

[0021] Preferably, step S1 includes the following steps: using a multi-branch state machine to determine the current assembly phase, which includes a synchronization waiting phase and a tracking phase; when it is determined that the current phase is a synchronization waiting phase, the product of the real-time virtual admittance coefficient and the relative synchronization deviation is determined as the micro position yield compensation amount, so that the heat sink core displacement adjustment mechanism generates a controlled elastic displacement.

[0022] Preferably, step S2 includes the following steps: performing a reduction-order operation on the acceleration state of the radiator core displacement adjustment mechanism, filtering out the second-order differential terms in the high-frequency noise components, and obtaining the disturbance estimate from the real-time spatial position feedback sequence as the input observation.

[0023] Preferably, step S5 includes the following steps: modulating the switching term gain in the sliding mode reaching law using the real-time virtual admittance coefficient, and increasing the system damping by increasing the real-time virtual admittance coefficient when the disturbance estimate exceeds the preset linear domain threshold, so as to suppress the transient chattering of the heat sink core displacement adjustment mechanism at the yield point of the heat sink fins.

[0024] Preferably, at the instant when the multi-branch state machine determines that the assembly phase has changed from the synchronous waiting phase to the tracking phase, the integral term coefficient in the control loop is adjusted according to the value of the disturbance estimate, and the historical accumulation path of the error is cut off by setting the integral term coefficient to zero.

[0025] Preferably, in step S2, the mechanical friction force is decoupled online using a reduced-order extended state observer, the contact force component reflecting the assembly interference intensity is separated from the disturbance estimate, and the contact force component is determined as the sole input source for the real-time virtual admittance coefficient calculation.

[0026] Preferably, the method utilizes real-time virtual admittance coefficients to establish dynamic matching rules between position adjustment amount and internal stress evolution of the core, generating controlled flexible displacement when the radiator core is closed.

[0027] Preferably, in step S3, the core stiffness parameters are preset according to the physical specifications of the heat sink to be assembled, and the response frequency matching the stiffness of the heat pipe array is obtained by adjusting the mapping slope between the real-time virtual admittance coefficient and the changing gradient.

[0028] Preferably, step S5 includes the following steps: sending the corrected convergence target position to the displacement adjustment driver, driving the displacement adjustment motor to drive the assembly displacement carrier to perform three-axis synchronous displacement compensation, and using the real-time virtual admittance coefficient to offset the relative synchronous deviation between multiple axes.

[0029] Example 1: In the automated assembly of radiator cores for high power density new energy vehicles, the heat exchanger tube array is pressed into a heat dissipation fin assembly. Due to the complex nonlinear elastoplastic mechanical characteristics of the heat dissipation fins, the yield resistance generated during multi-axis synchronous pressing exhibits irregular fluctuations with the increase of pressing depth. If a rigid position control command is used, in the multi-end synchronous waiting phase, the leading actuator remains locked in a fixed spatial position. At this time, the heat dissipation fins are compressed, generating elastic potential energy accumulation. As a result, at the instant when the synchronization error meets the requirements and the waiting state is released, the accumulated internal stress and the restored drive command superimpose, causing physical scale bounce and position overshoot, resulting in pressure damage to the precision heat dissipation fins.

[0030] In this scenario, the method of the present invention operates by extracting the real-time spatial position feedback sequence of multiple radiator core displacement adjustment mechanisms from the control unit, calculating the independent positional deviation of each radiator core displacement adjustment mechanism and the relative synchronization deviation between adjacent radiator core displacement adjustment mechanisms, and establishing a reduced-order extended state observer. By stripping the known linear dynamic terms from the nominal model of the system, a real-time numerical approximation is performed on the total lumped disturbance quantity composed of assembly contact resistance, mechanical friction, and clearance, and the transient disturbance compensation parameters are output. Through calculation The gradient of change is used to determine the real-time virtual admittance coefficient in conjunction with the preset core stiffness parameters. When the assembly enters the critical closing stage, and the multi-branch state machine determines that the independent position deviation of the specific radiator core displacement adjustment mechanism has dropped to within 0.05mm, and the maximum global synchronization error calculated from the encoder data exceeds 0.15mm, the system recognizes that it is currently in a synchronization waiting phase. At this time, it no longer fixes the position command, but instead calculates the transient disturbance compensation parameters. With real-time virtual admittance coefficient The product of these factors is used to determine the micro-position setback compensation amount. , Among them, represents the compensation amount for micro-positional setbacks. For real-time virtual admittance coefficients, This is a parameter for transient disturbance compensation.

[0031] The control unit uses micro-position yield compensation to adjust the convergence target position of the sliding mode approach law, driving the radiator core displacement adjustment mechanism to generate a controlled, compliant yield displacement that follows the fluctuations in assembly contact resistance. This action transforms the position command into dynamic energy tracking that conforms to the evolution of internal stress in the material. By allowing the leading actuator to generate micron-level adaptive yield when sensing significant resistance, it actively dissipates and releases the elastic potential energy accumulated inside the radiator core. As the lagging actuator completes displacement compensation, when the maximum global synchronization error decays to below 0.15mm, the system uses a dynamic admittance evaluation function to evaluate the real-time virtual admittance coefficient. Adjusting to 0 values ​​releases the virtual admittance adjustment logic, allowing the radiator core displacement adjustment mechanism to resume rigid position tracking towards the initial target. Since the internal stress has been physically released during the waiting phase, the system maintains a continuous motion trajectory during the transition from the synchronous waiting phase to the tracking phase, eliminating position jumps of the radiator core during state switching.

[0032] Example 2: This example demonstrates a test platform for verifying the matching degree between displacement adjustment commands and yield strength during radiator core assembly. The platform includes a multi-axis servo pressing mechanism and a displacement sensing device. The displacement sensing resolution in the test environment is 0.001. mm, sampling frequency set to 500 The sampling period (Hz) is set based on a balance between the real-time performance of data acquisition and the computational load of the processing unit. When the absolute value of the rate of change of assembly contact resistance exceeds the preset range, the sampling period tends to approach the lower limit of the range, i.e., 2Hz. ms, and the signal-to-noise ratio superimposed in the feedback sequence is 20. dB of background noise was used to simulate signal interference conditions in a metalworking workshop. In a comparative test during the heat sink core assembly stage, the control group used a fixed-position locking command. When the maximum global synchronization error reached 0.16... When the aluminum foil fins reach a yield strength of 42.5 N, the cumulative internal stress peak at the fin tip reaches 42.5 N. At the instant the synchronization condition is met and the locking is released, the position overshoot of the radiator core displacement adjustment mechanism is 0.12 mm. mm, the measured fin compression loss rate was 8.5%. For comparison, the sample group of this invention used a reduced-order extended state observer to extract transient perturbation compensation parameters from a sequence containing background noise. The system calculates the real-time virtual admittance coefficient based on preset core stiffness parameters. It is 0.015 mm / N, thereby determining the micro-position setback compensation amount. .

[0033] The radiator core displacement adjustment mechanism of the present invention responds The instruction generated 0.075 The adaptive compliant yielding displacement of mm dissipates the elastic potential energy accumulated inside the fin, reducing the measured peak internal stress at the fin tip to 9.2. N, as the lagging radiator core displacement adjustment mechanism catches up to an error of 0.15. Within mm, the system will Reduced to 0, position fluctuation controlled within 0.012. If the fin pressure loss rate is within mm and the fin pressure loss rate is 0%, in the control test exceeding the parameter limit range, if... Set to 0.05 If the yield is mm / N, excessive compliance leads to insufficient system damping and causes a frequency of 45.2. The forced oscillation of Hz, the data comparison results show that the adjustment range of the real-time virtual admittance coefficient takes into account both stress release and tracking accuracy; by establishing a dynamic mapping from transient disturbance compensation parameters to position variables, the method of this invention realizes the allocation of position adjustment weight between geometric tracking and energy balance in the processing environment with signal noise and nonlinear resistance fluctuations, and transforms position adjustment from size alignment to adaptive compliance in accordance with the mechanical characteristics of materials, so as to solve the problem of workpiece damage caused by the mismatch between rigid control and flexible material resistance in metal processing.

[0034] Example 3: This example combines Figures 1 to 2 This paper describes a high-precision position adjustment method in the assembly of a radiator core, such as... Figure 1 As shown, the high-precision position adjustment method in the assembly of the radiator core includes step S1 for acquiring the real-time spatial position feedback sequence of multiple radiator core displacement adjustment mechanisms, and calculating the independent position deviation of each radiator core displacement adjustment mechanism and the relative synchronization deviation between adjacent radiator core displacement adjustment mechanisms; step S2 for establishing a reduced-order expansion state observer, and extracting the estimated quantity reflecting the assembly contact resistance disturbance based on the real-time spatial position feedback sequence, wherein the reduced-order expansion state observer uses the system nominal model to obtain the acceleration term deviation generated by the external resistance; and step S3 for acquiring the gradient of the disturbance estimate change, and determining the core stiffness parameter based on the gradient change and the preset core stiffness parameter. The real-time virtual admittance coefficient includes step S4, which determines that when the independent position deviation of a specific radiator core displacement adjustment mechanism drops to within 0.05mm and the maximum global synchronization error exceeds 0.15mm, the product of the disturbance estimate and the real-time virtual admittance coefficient is calculated, and the calculation result is determined as the micro-position yield compensation amount. Step S5 is used to adjust the sliding mode convergence law to the target position using the micro-position yield compensation amount, driving the radiator core displacement adjustment mechanism to generate a displacement following the assembly contact resistance fluctuation until the maximum global synchronization error decays to below 0.15mm, and then the real-time virtual admittance coefficient is adjusted to zero to resume tracking towards the initially set target.

[0035] like Figure 2As shown, this is achieved through the interaction between the real-time control unit and the field execution nodes. The real-time control unit includes a reduced-order extended state observer, a multi-branch state machine, a dynamic admittance evaluation module, and a computation module consisting of a micro-position setback compensation calculation, sliding mode convergence target correction, and integral term coefficient adjustment. The reduced-order extended state observer constructs and performs acceleration term deviation acquisition and disturbance estimate extraction based on the system nominal model. The multi-branch state machine performs assembly phase determination, synchronous waiting phase identification, and tracking phase identification and outputs phase trigger signals. The dynamic admittance evaluation module performs gradient calculation and real-time... The virtual admittance coefficient is determined and the core stiffness parameters are matched, and the real-time virtual admittance coefficient is output. The aforementioned disturbance estimate, phase trigger signal and real-time virtual admittance coefficient are gathered in the calculation module to generate the corrected convergence target position command. The field execution node includes a displacement sensing device and a displacement adjustment mechanism group. The displacement sensing device includes position encoder feedback and spatial position sequence acquisition function. The acquired real-time spatial position feedback sequence is sent to the real-time control unit. The displacement adjustment mechanism group includes a displacement adjustment motor and an assembly displacement carrier, which is used to receive the corrected convergence target position command and perform triaxial synchronous displacement compensation.

[0036] Example 4: In a metalworking scenario for assembling a heavy-duty industrial heat exchanger core, due to the use of aluminum alloy foils of varying thicknesses for the heat sink fins, the material stiffness properties of different batches deviate. Before the assembly cycle begins, the control unit initiates static stiffness detection, driving the heat sink core displacement adjustment mechanism to press against the reference block at a speed of 1 mm / s. Simultaneously, it acquires the stroke increment of the displacement sensor and the initial disturbance value fed back by the expansion state observer. When the disturbance estimate changes from 0 to 5 N and remains stable for 50 ms, the displacement change is recorded as 0.1 mm. Based on the linear proportional relationship between force and displacement, the system determines the preset core stiffness parameter as 50 N / mm. For the discretized logic of the reduced-order expansion state observer, the control unit constructs a system state update equation. Within each 2 ms control cycle, the system reads the position components in the real-time spatial position feedback sequence. And utilize the control output from the previous moment Update the observer's internal state and observer gain. Based on the system bandwidth, set the expected bandwidth of the observer. 250 rad / s, gain Set to 250, the observer uses the collected acceleration term deviation to fit the dynamic evolution law of the external assembly contact resistance, thus making the transient disturbance compensation parameter... To maintain a resolution of 0.005N under signal interference, when extracting the acceleration term deviation, in order to avoid directly performing a second difference on the discrete spatial position feedback sequence, which would cause high-frequency environmental noise to be amplified quadratically, the order reduction operation process is performed by calling the discrete tracking differentiator module inside the control unit to perform a first-order smoothing filter on the real-time spatial position feedback sequence to obtain its lossless velocity estimation component. Based on the inherent dynamic equations of the nominal model, the system performs a feedforward subtraction operation on the first-order velocity estimation component and the current motor servo torque output command, thereby reconstructing the true acceleration state through analytical substitution without calling any second-order differential operators, thus cutting off the high-frequency amplification path of noise at the bottom layer of the algorithm architecture.

[0037] The system calculates The first-order difference is used to obtain the gradient of the change in the perturbation estimator. and establish by To real-time virtual admittance coefficient The mapping relationship is such that when the independent position deviation of a specific radiator core displacement adjustment mechanism drops to 0.05mm and the maximum global synchronization error exceeds 0.15mm, if the actual measurement... The value is 150 N / s, and the system will adjust the scaling factor accordingly. The value was determined to be 0.015 mm / N, and then the micro-position setback compensation was calculated. : ,in, This refers to the compensation amount for the micro-position setback. For real-time virtual admittance coefficients, For transient disturbance compensation parameters, during the aforementioned numerical conversion, the dynamic admittance evaluation function configured within the control unit is specifically constructed as a first-order linear mapping operator. The scaling factor in this operator is obtained by multiplying the inverse of the preset core stiffness parameter by a constant safety margin coefficient. The real-time virtual admittance coefficient is the direct product of the change gradient of the disturbance estimate and this scaling factor. Through this explicit algebraic relationship locking, the system ensures that the calculated output flexible damping parameter under different resistance abrupt gradients exhibits a proportional evolution trajectory with a definite slope, preventing state solution divergence. Since the radiator core displacement adjustment mechanism generates a compliant yield displacement of 0.075mm instantaneously upon sensing an increase in contact resistance, this action corrects the convergence target of the sliding mode approach law from zero to the stress release equilibrium point, realizing the dissipation of assembly internal stress. As the lagging actuator completes the synchronization error compensation and the maximum global synchronization error decays to below 0.15mm, the system uses a decreasing function to... Within 20ms, the value is reduced to 0, allowing the radiator core displacement adjustment mechanism to resume tracking the target, completing the closed-loop process from physical force perception to geometric precision alignment.

[0038] Example 5: In the deployment scenario of a multi-specification heat exchanger hybrid production line, for heat dissipation fins made of aluminum-manganese alloy or composite brazed foil, the system generates a correlation data sequence of displacement and disturbance before the formal assembly cycle starts. The control unit drives the heat dissipation core adjustment mechanism to move to the surface of the core to be assembled in increments of 1 mm / s. The reduced-order expansion state observer acquires the acceleration term deviation caused by physical contact in real time and determines the transient disturbance compensation parameters. The system records the correspondence between the displacement change and the disturbance estimate using discrete numerical values. It then determines the tangent slope of the heat sink fins at the critical point of elastic-plastic transition using the least squares method. This slope is stored in the control unit as a preset core stiffness parameter. This allows the system to determine the real-time virtual admittance coefficient online based on the obtained slope when dealing with different batches of raw materials with varying physical properties. While determining the parameters through slope fitting, the system processor locks the extreme values ​​of the disturbance contact force on the physical coordinates of the aforementioned elastoplastic transition critical point, and writes the mechanical absolute value corresponding to the yield stress starting point into memory as a preset linear domain threshold. When the transient disturbance calculated in the actual assembly process exceeds the threshold limit, it indicates that the workpiece has left the pure elastic response state. This threshold extraction scheme based on real-time sampling and in-situ calibration ensures that the control intervention node can accurately track the substantial changes in the workpiece deformation state.

[0039] When the system is in an assembly phase where electromagnetic noise causes fluctuations in the position feedback signal, the control unit determines the judgment threshold by statistically analyzing the variance of the feedback deviation of the radiator core displacement adjustment mechanism during the static positioning process. The system collects 1000 position sample points during the reciprocating stroke and generates a dynamic deviation envelope of the position feedback sequence. 1.2 times the peak value of the envelope is determined as the independent position deviation threshold of 0.05mm for determining that the radiator core displacement adjustment mechanism has entered the controlled compliant yielding phase. Simultaneously, the boundary of the global synchronization error maximum value is determined to be 0.15mm by analyzing the synchronization error frequency of adjacent actuators during cooperative acceleration. This value serves as the trigger criterion for correcting the target position of the sliding mode convergence law. When the system detects that the global synchronization error maximum value has decayed below the boundary, the system uses a decreasing function... The correction amount of the real-time virtual admittance coefficient is reduced, and the heat sink core displacement adjustment mechanism returns from a compliant yielding state to a tracking state toward the set target. In the above-mentioned cross-scale parameter mapping process, the servo torque pulsation of adjacent actuators during coordinated acceleration will excite mechanical resonance of characteristic frequency. The reciprocal of the synchronization error frequency directly corresponds to the forced vibration period of the mechanical system in the physical domain. The control unit uses a discrete time domain integration algorithm to accumulate and evolve the maximum following speed residual sequence within the vibration period time window, accurately converting the energy characteristics of high-frequency oscillation into the transient deviation displacement peak value in three-dimensional spatial geometry. Using this displacement extreme value as the theoretical deduction lower limit, the physical boundary of the global synchronization error of 0.15mm is derived and determined, so that the frequency characteristics and spatial distance have a self-consistent conversion channel.

[0040] Example 6: In the on-site commissioning of a high-speed new energy radiator core assembly line, the control unit initializes the state variables of the radiator core displacement adjustment mechanism. During the power-on startup phase, the system resets the state variables of each order of the reduced-order extended state observer to 0 values ​​and reads the real-time spatial position feedback sequence of the displacement adjustment mechanism under the rapid advance phase. When the multi-branch state machine detects that the independent position deviation of a specific radiator core displacement adjustment mechanism has dropped to 0.05mm and the maximum global synchronization error calculated using encoder feedback exceeds 0.15mm, the logic judgment unit controls the system to enter the energy... In the dissipation yield mode, to adapt to the size differences of workpieces with different physical specifications processed by the mixed production line, the system control architecture has a built-in dynamic scale scaling correction item. The 0.05mm and 0.15mm in the above criteria are used as the basic configuration reference values. Before the actual working condition is called, the control unit will extract the scale scaling coefficient according to the average mechanical spacing characteristics of the heat sink of the new generation heat exchanger core. After performing a multiplication operation between the basic reference value and the scaling coefficient, it will be sent out for execution. In this way, the fixed judgment threshold boundary can be adaptively generalized and evolved according to the structural characteristics of the object to be assembled, overcoming the limitations caused by specific test materials.

[0041] Within the control cycle of the micro-position retraction action of the radiator core displacement adjustment mechanism, the convergence target of the sliding mode approach law is adjusted from zero to the micro-position retraction compensation amount. The locked offset point is compensated by the transient disturbance parameters output by the observer. With real-time virtual admittance coefficient Calculate the dynamic switching gain of the sliding surface and lock the switching gain to [value missing]. The absolute value is 1.1 times. The corrected reference target point is set to be equal to the initial issued spatial tracking target minus the micro-yield compensation amount. The current working condition position tracking deviation is calculated. The preset sliding surface slope parameter is introduced to construct a linear switching surface. Then, according to the equivalent approximation law, the motor drive torque command is continuously output based on the first-order difference of the position deviation, the second-order difference of the reference trajectory, and the transient disturbance compensation parameter. The sign function directly flips the compensation output direction according to the positive and negative polarities of the phase trajectory where the sliding surface is located. The drive servo circuit physically follows the bottom nonlinear resistance trough to produce yielding. During the process of the controlled and compliant yielding displacement of the specific radiator core displacement adjustment mechanism, the control unit compares the position tracking residual. When the maximum value of the global synchronization error decays to below 0.15mm, the system stops the admittance correction logic and resets the controller integral term. The radiator core displacement adjustment mechanism resumes the tracking phase towards the initially set target.

[0042] In operating conditions accompanied by high-frequency mechanical vibration and requiring the maintenance of synchronization accuracy, the control unit will compensate for the micro-position yield. In the deviation calculation stage of the sliding mode approach law, the system calculates the initial target position and the micro-position back-off compensation amount. The difference is used to generate a corrected reference position. The sliding mode control law calculates the control output in real time based on the deviation between the corrected reference position and the real-time feedback position of the radiator core displacement adjustment mechanism. This process enables the radiator core displacement adjustment mechanism to maintain the convergence characteristics of the sliding surface during the period of generating controlled compliant yield displacement, eliminating the actuator drive current pulse caused by the jump of the position command target value. When the system detects that the maximum value of the global synchronization error has decayed to below 0.15mm and triggers the state switching logic, the logic judgment unit controls the real-time virtual admittance coefficient through a linear proportional decay function. During the removal process, the linear proportional decay function uniformly reduces the correction weight of the virtual admittance from 1.0 to 0 within a 20ms time window. The control unit uses the reduced weight to correct the micro position yield compensation amount, so that the command torque of the radiator core displacement adjustment mechanism remains numerically continuous when switching from the energy dissipation yield state to the rigid tracking state.

[0043] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A high-precision position adjustment method in the assembly of a radiator core, characterized in that, Includes the following steps: Step S1: Obtain the real-time spatial position feedback sequence of multiple radiator core displacement adjustment mechanisms, and calculate the independent position deviation of each radiator core displacement adjustment mechanism and the relative synchronization deviation between adjacent radiator core displacement adjustment mechanisms. Step S2: Establish a reduced-order extended state observer and extract the disturbance estimate reflecting the assembly contact resistance based on the real-time spatial position feedback sequence. The reduced-order extended state observer uses the system nominal model to obtain the acceleration term deviation caused by the external resistance. Step S3: Obtain the gradient of the change in the disturbance estimate, and determine the real-time virtual admittance coefficient based on the gradient and the preset core stiffness parameters. Step S4: When the independent position deviation of the specific radiator core displacement adjustment mechanism is reduced to within 0.05mm and the maximum global synchronization error exceeds 0.15mm, calculate the product of the disturbance estimate and the real-time virtual admittance coefficient, and determine the calculation result as the micro position yield compensation amount. Step S5: Adjust the convergence target position of the sliding mode approach law by using the micro position yield compensation amount, drive the heat sink core displacement adjustment mechanism to generate displacement that follows the fluctuation of assembly contact resistance, until the maximum value of global synchronization error decays to below 0.15mm, then adjust the real-time virtual admittance coefficient to zero value to restore tracking to the initially set target.

2. The high-precision position adjustment method in assembling a radiator core according to claim 1, characterized in that, Step S1 includes the following steps: using a multi-branch state machine to determine the current assembly phase, which includes a synchronization waiting phase and a tracking phase; when it is determined that the current phase is a synchronization waiting phase, the product of the real-time virtual admittance coefficient and the relative synchronization deviation is determined as the micro position yield compensation amount, so that the heat sink core displacement adjustment mechanism generates a controlled elastic displacement.

3. The high-precision position adjustment method in assembling a radiator core according to claim 1, characterized in that, Step S2 includes the following steps: performing a reduction-order operation on the acceleration state of the radiator core displacement adjustment mechanism, filtering out the second-order differential terms in the high-frequency noise components, and obtaining the disturbance estimate from the real-time spatial position feedback sequence as the input observation.

4. The high-precision position adjustment method in assembling a radiator core according to claim 1, characterized in that, Step S5 includes the following steps: modulate the switching term gain in the sliding mode reaching law using the real-time virtual admittance coefficient. When the disturbance estimate exceeds the preset linear domain threshold, increase the system damping by increasing the real-time virtual admittance coefficient to suppress the transient chattering of the heat sink core displacement adjustment mechanism at the heat sink fin yield point.

5. A high-precision position adjustment method for assembling a radiator core according to claim 2, characterized in that, When the multi-branch state machine determines that the assembly phase changes from the synchronous waiting phase to the tracking phase, the integral term coefficient in the control loop is adjusted according to the value of the disturbance estimate. The historical accumulation path of the error is cut off by setting the integral term coefficient to zero.

6. A high-precision position adjustment method for assembling a radiator core according to claim 1, characterized in that, In step S2, the mechanical friction force is decoupled online using a reduced-order extended state observer, the contact force component reflecting the assembly interference intensity is separated from the disturbance estimate, and the contact force component is determined as the sole input source for the calculation of the real-time virtual admittance coefficient.

7. A high-precision position adjustment method for assembling a radiator core according to claim 1, characterized in that, The method utilizes real-time virtual admittance coefficients to establish dynamic matching rules between position adjustment and internal stress evolution of the core, generating controlled flexible displacement when the radiator core is closed.

8. A high-precision position adjustment method for assembling a radiator core according to claim 1, characterized in that, In step S3, the core stiffness parameters are preset according to the physical specifications of the heat sink to be assembled. By adjusting the mapping slope between the real-time virtual admittance coefficient and the changing gradient, the response frequency that matches the stiffness of the heat pipe array is obtained.

9. A high-precision position adjustment method for assembling a radiator core according to claim 1, characterized in that, Step S5 includes the following steps: sending the corrected convergence target position to the displacement adjustment driver, driving the displacement adjustment motor to drive the assembly displacement carrier to perform three-axis synchronous displacement compensation, and using the real-time virtual admittance coefficient to offset the relative synchronous deviation between multiple axes.

Images