Intelligent assembly monitoring management system

Abstract

The application relates to the field of industrial control systems and discloses an intelligent assembly monitoring management system, which comprises a state acquisition module, a transient power flow analysis module and an active damping injection module arranged in a servo driving unit; the state acquisition module acquires a stator q-axis current feedback value and a rotor instantaneous angular velocity of a controlled object; the transient power flow analysis module extracts current and angular velocity pulsation components by using a high-pass filter and generates a transient active power scalar according to the product; and the active damping injection module generates a reverse damping current instruction and injects the current control loop when the scalar polarity is reversed; the application captures the transient evolution of the electromechanical energy conversion direction, establishes the physical boundary for distinguishing normal friction dissipation and abnormal collision feedback, realizes in-situ defense by bypassing the polling cycle of the main controller, and suppresses physical damage caused by assembly interference from the bottom logic level.

Description

Technical Field

[0001] This invention relates to an intelligent assembly monitoring and management system, belonging to the field of industrial control system technology. Background Technology

[0002] The current mainstream control architecture adopts a closed-loop regulation method that separates the host controller and servo driver. It uses an industrial bus to periodically collect sensor data and schedules the assembly status through the logic layer of the host controller. This structure has good process versatility in conventional manufacturing conditions. However, in the assembly process of aerospace-grade precision components, the interference evolution of the assembly interface is on the order of microseconds. The command scheduling link of the master-slave architecture has an unavoidable time delay. The price of maintaining control stability in the current method is the sacrifice of the ability to intercept transient overload stress. When unexpected interference occurs at the assembly interface, the lag in the monitoring cycle causes the driver to be unable to shut off the torque output before physical damage occurs, which in turn causes component lattice damage or stress concentration.

[0003] To shorten response time, increasing the bus sampling frequency or lowering the judgment threshold are intuitive improvement paths. However, simply increasing the frequency will introduce high-frequency vibration and electromagnetic interference, causing unexpected oscillations in the control loop. At the same time, the judgment method based on torque amplitude cannot distinguish between normal tight-fit friction and abnormal collision stress. Existing solutions have limitations in the dimension of control strategy feature recognition. For example, Chinese invention patent CN109202907B discloses a sensorless elastic collision device and method based on industrial robots. It obtains velocity and current center curves through offline learning and judges collisions based on deviations crossing thresholds. This detection logic is a macroscopic collision stress triggering mechanism. The judgment benchmark relies on the static mapping of the envelope, which makes it unable to penetrate complex tight-fit friction noise and difficult to identify transient changes in the direction of microsecond-level electromechanical energy conversion. In precision assembly, the microscopic evolution of assembly gaps is easily submerged in steady-state load deviation fluctuations, which leads to judgment lag or malfunctions near the physical damage threshold.

[0004] Therefore, how to achieve feature extraction and transient suppression of assembly interference within the servo drive closed loop is 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: An intelligent assembly monitoring and management system, comprising a main controller and a servo drive unit: The servo drive unit is internally equipped with a mutually coupled state acquisition module, transient power flow analysis module, and active damping injection module; The status acquisition module is used to acquire the stator q-axis current feedback value and rotor position feedback value of the controlled object, and to perform differential calculation using the position deviation between adjacent sampling periods to obtain the instantaneous angular velocity of the rotor; The transient power flow analysis module is equipped with a high-pass filter with a cutoff frequency of 500Hz, which is used to strip the DC component from the stator q-axis current feedback value and the rotor instantaneous angular velocity to obtain the q-axis current pulsation component and the angular velocity pulsation component. The transient power flow analysis module is also equipped with a multiply-accumulate operator, which is used to discretely integrate the product of the q-axis current pulsation component and the angular velocity pulsation component within a sliding time window of 100μs to 500μs to generate an active power scalar that characterizes the direction of electromechanical energy conversion of the controlled object. The active damping injection module is connected to the transient power flow analysis module. It is used to read the positive and negative polarities of the active power scalar. When the active power scalar reverses from a positive value to a negative value, it determines that the controlled object has entered the mechanical energy feedback state, obtains the absolute value of the negative value, and proportionally maps it with the preset electromagnetic dissipation coefficient to generate a reverse damping current command. The servo drive unit is used to bypass the polling cycle of the main controller and inject the reverse damping current command as a feedforward compensation parameter into the stator q-axis current control loop of the servo drive unit to generate a braking torque opposite to the rotor rotation direction and dissipate the mechanical rebound energy generated by the controlled object.

[0006] Preferably, the status acquisition module is equipped with a high-frequency sampling clock, the sampling frequency of which is not less than 10kHz. The status acquisition module uses the high-frequency sampling clock to sample the rotor position feedback value at equal intervals to obtain a discrete position sequence, and calculates the ratio of the position deviation between two adjacent sampling periods in the discrete position sequence to the reciprocal of the sampling frequency to determine the rotor instantaneous angular velocity. The rotor instantaneous angular velocity and the stator q-axis current feedback value are kept in phase synchronization on the time axis to eliminate the time-domain phase deviation when generating the active power scalar.

[0007] Preferably, the high-pass filter is a second-order infinite impulse response filter; the transient power flow analysis module uses the high-pass filter to dynamically truncate the stator q-axis current feedback value, stripping away the low-frequency components that characterize the constant operating resistance of the controlled object, and locking in the pulsating characteristics that characterize the transient physical interference of the controlled object.

[0008] Preferably, the transient power flow analysis module is equipped with a sliding shift register, which is used to store discrete data sequences within a sliding time window; whenever the status acquisition module generates new sampling data, the sliding shift register removes the earliest set of data and pushes in the latest set of data, thus completing the real-time sliding update of the active power scalar.

[0009] Preferably, the active damping injection module stores a preset negative power threshold. When the active power scalar is less than the negative power threshold, the active damping injection module determines that the controlled object has undergone a rigid collision and sends a reverse torque signal to the current loop controller of the servo drive unit.

[0010] Preferably, the servo drive unit is also equipped with an abnormal alarm module, which is used to upload interference characteristic data characterizing the real-time operating condition of the controlled object to the main controller at the same time as the active damping injection module generates the reverse damping current command, so as to cause the main controller to interrupt the current controlled task sequence.

[0011] Preferably, the negative power threshold is set based on the power fluctuation variance of the system under no-load conditions; the active damping injection module performs statistical sampling on the active power scalar under no-load conditions and sets three times the power fluctuation variance as the judgment boundary of the negative power threshold, which is used to shield the random electromagnetic noise inside the servo drive unit.

[0012] Preferably, the servo drive unit is a multi-axis integrated driver, and the status acquisition module simultaneously acquires the current feedback and position feedback of multiple motion axes; the transient power flow analysis module calculates the active power scalar corresponding to each motion axis respectively, and when the polarity of any motion axis is reversed, the active damping injection module locks all related controlled motion axes in a coordinated manner.

[0013] Preferably, the servo drive unit is further provided with an impedance mapping module; the impedance mapping module calculates the transient impedance characteristics of the controlled object based on the slope of the change of the active power scalar, and maps the transient impedance characteristics to the feedback gain of the active damping injection module, so as to realize compliant protection of the physical properties of different controlled objects.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. In intelligent assembly monitoring and management, a state recognition mechanism based on the polarity of energy flow is constructed to establish the physical boundary that distinguishes mechanical contact friction loss from rigid collision interference. During normal precision assembly, the servo motor maintains positive active power output to overcome guiding friction. When micro-interference causes transient rebound of mechanical potential energy, this energy feedback process drives the transient active power to cross zero and reverse to a negative value. Based on the polarity change of the power flow, the system directly triggers active damping commands at the bottom layer, solving the problem of false shutdown when relying on torque amplitude for judgment in traditional solutions.

[0015] 2. This invention achieves the co-location of assembly monitoring logic and servo current closed loop, eliminating the communication delay bottleneck in the control architecture. By embedding a transient impedance mapping module inside the servo drive unit, the torque acquisition signal can directly participate in the real-time adjustment of the underlying control law, bypassing the command issuance link of the upper controller. This mechanism enables the system to intercept destructive torque before stress concentration is caused by physical contact, improving the dynamic response capability of the control system to microsecond-level transient disturbances.

[0016] 3. This invention utilizes a multi-axis kinematic Jacobian matrix to construct a cross-decoupled compliant control mode, solving the problem of rigid potential energy accumulation under complex working conditions. When a specific motion axis triggers an interference warning, the control loop synchronously adjusts the servo position loop stiffness parameters in the orthogonal direction according to the Jacobian mapping coefficient, inducing the actuator to generate physical compensation displacement along the non-force direction. This multi-feature coordinated adjustment strategy enables the system to have adaptive obstacle avoidance tolerance, preventing lattice damage or micro-crushing of high-precision components at the moment of jamming. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the energy flow direction determination and in-situ defense process of the intelligent assembly monitoring and management system of the present invention. Figure 2 This diagram shows the multi-level hardware nodes and functional modules of the intelligent assembly monitoring and management system 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] An intelligent assembly monitoring and management system includes a main controller and a servo drive unit. The servo drive unit is internally equipped with a mutually coupled state acquisition module, transient power flow analysis module, and active damping injection module; The status acquisition module is used to acquire the stator q-axis current feedback value and rotor position feedback value of the controlled object, and to perform differential calculation using the position deviation between adjacent sampling periods to obtain the instantaneous angular velocity of the rotor; The transient power flow analysis module is equipped with a high-pass filter with a cutoff frequency of 500Hz, which is used to strip the DC component from the stator q-axis current feedback value and the rotor instantaneous angular velocity to obtain the q-axis current pulsation component and the angular velocity pulsation component. The transient power flow analysis module is also equipped with a multiply-accumulate operator, which is used to discretely integrate the product of the q-axis current pulsation component and the angular velocity pulsation component within a sliding time window of 100μs to 500μs to generate an active power scalar that characterizes the direction of electromechanical energy conversion of the controlled object. The active damping injection module is connected to the transient power flow analysis module. It is used to read the positive and negative polarities of the active power scalar. When the active power scalar reverses from a positive value to a negative value, it determines that the controlled object has entered the mechanical energy feedback state, obtains the absolute value of the negative value, and proportionally maps it with the preset electromagnetic dissipation coefficient to generate a reverse damping current command. The servo drive unit is used to bypass the polling cycle of the main controller and inject the reverse damping current command as a feedforward compensation parameter into the stator q-axis current control loop of the servo drive unit to generate a braking torque opposite to the rotor rotation direction and dissipate the mechanical rebound energy generated by the controlled object.

[0021] Preferably, the status acquisition module is equipped with a high-frequency sampling clock, the sampling frequency of which is not less than 10kHz. The status acquisition module uses the high-frequency sampling clock to sample the rotor position feedback value at equal intervals to obtain a discrete position sequence, and calculates the ratio of the position deviation between two adjacent sampling periods in the discrete position sequence to the reciprocal of the sampling frequency to determine the rotor instantaneous angular velocity. The rotor instantaneous angular velocity and the stator q-axis current feedback value are kept in phase synchronization on the time axis to eliminate the time-domain phase deviation when generating the active power scalar.

[0022] Preferably, the high-pass filter is a second-order infinite impulse response filter; the transient power flow analysis module uses the high-pass filter to dynamically truncate the stator q-axis current feedback value, stripping away the low-frequency components that characterize the constant operating resistance of the controlled object, and locking in the pulsating characteristics that characterize the transient physical interference of the controlled object.

[0023] Preferably, the transient power flow analysis module is equipped with a sliding shift register, which is used to store discrete data sequences within a sliding time window; whenever the status acquisition module generates new sampling data, the sliding shift register removes the earliest set of data and pushes in the latest set of data, thus completing the real-time sliding update of the active power scalar.

[0024] Preferably, the active damping injection module stores a preset negative power threshold. When the active power scalar is less than the negative power threshold, the active damping injection module determines that the controlled object has undergone a rigid collision and sends a reverse torque signal to the current loop controller of the servo drive unit.

[0025] Preferably, the servo drive unit is also equipped with an abnormal alarm module, which is used to upload interference characteristic data characterizing the real-time operating condition of the controlled object to the main controller at the same time as the active damping injection module generates the reverse damping current command, so as to cause the main controller to interrupt the current controlled task sequence.

[0026] Preferably, the negative power threshold is set based on the power fluctuation variance of the system under no-load conditions; the active damping injection module performs statistical sampling on the active power scalar under no-load conditions and sets three times the power fluctuation variance as the judgment boundary of the negative power threshold, which is used to shield the random electromagnetic noise inside the servo drive unit.

[0027] Preferably, the servo drive unit is a multi-axis integrated driver, and the status acquisition module simultaneously acquires the current feedback and position feedback of multiple motion axes; the transient power flow analysis module calculates the active power scalar corresponding to each motion axis respectively, and when the polarity of any motion axis is reversed, the active damping injection module locks all related controlled motion axes in a coordinated manner.

[0028] Preferably, the servo drive unit is further provided with an impedance mapping module; the impedance mapping module calculates the transient impedance characteristics of the controlled object based on the slope of the change of the active power scalar, and maps the transient impedance characteristics to the feedback gain of the active damping injection module, so as to realize compliant protection of the physical properties of different controlled objects.

[0029] Example 1: In automated assembly lines involving precision electromechanical components, the assembly interface generates unsteady frictional resistance due to tight fit interference. The torque signal generated by this resistance is highly similar in amplitude characteristics to the stress generated by mechanical jamming. This makes traditional monitoring architectures based on torque threshold judgment unable to distinguish between normal frictional work and unexpected mechanical interference. If the judgment threshold is lowered, the servo system will frequently generate false alarms and stop. If the threshold is maintained at a high level, the communication delay of the main controller will prevent the system from shutting off the torque output before physical damage to the components occurs. The operation mode of this technical solution under this condition is as follows: the status acquisition module inside the servo drive unit samples at 10kHz. The frequency acquisition module obtains the actual feedback values ​​of the stator q-axis current and rotor position of the controlled object, and performs high-frequency differential calculation on the actual rotor position feedback value to obtain the instantaneous rotor angular velocity. At this time, the high-pass filter inside the transient power flow analysis module sets its cutoff frequency to 500Hz to filter out the low-frequency DC steady-state components in the actual stator q-axis current feedback value and the instantaneous rotor angular velocity, thereby obtaining the q-axis current pulsation component and the angular velocity pulsation component. The multiply-accumulate operator inside the transient power flow analysis module performs discrete integration on the product of the above pulsation components within a 100μs sliding time window to generate the transient active power scalar, the specific calculation formula of which is as follows: ,in, For transient active power scalar, For the first The q-axis current ripple component at each sampling time. For the first Angular velocity pulsation components at each sampling time In this embodiment, the total number of sampling points within a 100μs sliding time window is [the number of sampling points]. It should be noted that the value of the total number of sampling points N is set to 1, based on the sliding time window. With system sampling period The ratio is adaptively determined. In this embodiment, at a sampling frequency of 10kHz, for a minimum window of 100μs... , When the value is 1, the calculation result reflects the instantaneous power polarity of a single sample; however, when the assembly process switches to a micro-interference detection scenario requiring a higher signal-to-noise ratio, the sliding time window... When configured to 500μs, the system will automatically push in 5 consecutive sets of pulsating components using a sliding shift register, and perform discrete summation and integration of the product of the current time and the past 4 sampling periods. Through this dynamic adjustment of the window width, the multiply-accumulate unit can use the low-pass filtering characteristics of discrete integration to suppress random sampling noise, thereby extracting stable energy feedback characteristics in assembly interfaces of different complexities.

[0030] Based on the law of conservation of electromechanical energy conversion, a preset torque constant of the servo motor is input before the scalar output of the multiply-accumulate operator. Perform dimensionless conversion and preset torque constant. The electromagnetic torque generated by the unit q-axis current of the servo motor, with the international standard unit being N·m / A, is pre-written and stored in the driver's read-only register, enabling the acquisition of transient active power scalar values. Possessing objective mechanical power physical dimensions, within the closed-loop framework of the synchronous rotating coordinate system of the servo drive unit, since the system maintains zero d-axis current and the stator winding inductive reactance exhibits a linear relationship at the pulsating frequency, the q-axis current component directly characterizes the evolution of the system's total active current. This is due to the back electromotive force generated by the motor and the rotor angular velocity and torque constant. The product is proportional; therefore, directly multiplying the q-axis current ripple after removing the DC component with the angular velocity ripple is physically equivalent to extracting the product of the active voltage term and the current term related to torque work in the motor terminal voltage. This calculation logic eliminates the steady-state voltage drop fluctuations caused by motor leakage inductance and armature resistance at the algorithm's underlying level, thereby... It can purely extract the transient mechanical work interaction characteristics of the electromechanical conversion interface. Based on this calculation logic, the impedance mapping module extracts the scalar difference of active power within adjacent time windows to quantify the transient power change rate, according to the viscoelastic damping physical model. The impedance mapping module uses a linear mapping formula. Determine the feedback gain To ensure the dimensional rigor of control commands during physical domain conversion, feedback gain is used. The mapping logic implicitly includes units of The conductance mapping operator is derived in real time from the reciprocal of the back electromotive force constant of the servo motor at the current operating speed. Its physical essence is to calculate and obtain the transient active power feedback quantity. This is converted into the corresponding equivalent electromagnetic damping current amplitude; therefore, the impedance mapping module calculates and obtains... The result of multiplying by the absolute value of the negative power scalar directly corresponds to the current vector that can generate braking torque within the current loop, thus eliminating the dimensional deviation when crossing from the power domain to the current domain. The basic damping constant obtained by calibration under the no-load steady-state operation of the system is set as a dimensionless parameter. Given the slope coefficient of the scale, The impedance mapping module calculates the feedback gain based on the scalar difference of active power between the current sampling period and the previous sampling period. Subsequently, the active power scalar is multiplied by the absolute value of its negative value when the polarity reverses, generating a reverse damping current command with dynamic amplitude that matches the current physical rebound state of the controlled object. The active damping injection module continuously reads this scalar. When the transient active power scalar remains positive, it is determined that the system is in a positive work state against frictional torque, maintaining the output setting of the original position control loop. When mechanical interference occurs at the assembly end and triggers potential energy rebound, causing the transient active power scalar to reverse to a negative value, the active damping injection module obtains the absolute value of this scalar in real time and generates a reverse damping current command, which is injected into the stator q-axis current as a compensation parameter. Within the flow control loop, the above process transforms the traditional torque amplitude discrimination into state recognition based on the polarity of energy flow. The high-pass filter strips the DC component, enabling the pulsating signal to reflect the microsecond-level torque fluctuation characteristics. Meanwhile, the transient power scalar provided by the multiply-accumulate unit establishes a physical boundary that distinguishes between frictional losses and potential energy feedback. This polarity determination logic, executed in situ within the servo drive unit, allows the system to actively absorb the rebound energy generated by assembly interference using reverse damping current, while avoiding the communication delay of the main controller bus. This enables dynamic suppression of unexpected stress before the physical contact force exceeds the component damage threshold.

[0031] Example 2: In a physical test environment simulating the assembly of precision electromechanical components, the test platform includes a six-degree-of-freedom assembly robot and a matching servo control cabinet. The current loop sampling frequency of the servo drive unit is set to 10kHz, and the position feedback resolution is set to 20bit. To construct an electromagnetic interference environment consistent with industrial settings, a Gaussian white noise signal with a signal-to-noise ratio of 20dB is superimposed on the signal transmission link. A pressure sensor with a measured range of 0 to 100N and a resolution of 0.01N is used as a reference. The cutoff frequency of the high-pass filter is... It is a core variable, and its setting involves a trade-off between filtering out steady-state load components and retaining transient interference characteristics. Below 200Hz, the low-frequency component of residual frictional torque in the signal causes the power scalar polarity determination to fail, while when At frequencies above 800Hz, the phase lag generated by the filter causes the damping injection compensation time to deviate from the stress peak value. Therefore, 500Hz was selected as the preferred cutoff frequency in this experiment.

[0032] The test process was divided into three gradients: standard assembly condition, slight misalignment interference condition, and severe collision condition. Under the standard assembly condition, the actual feedback value of the stator q-axis current obtained by the state acquisition module fluctuated around 1.2A. After processing by the transient power flow analysis module, the transient active power scalar was generated. Maintaining a power level between 0.5W and 1.8W indicates that the energy flow is from the power grid to the mechanical load, and the system maintains the predetermined position control command. When entering a micro-misalignment interference condition, mechanical oscillation occurs at the assembly interface, and the q-axis current pulsation component extracted by the transient power flow analysis module is observed. A fluctuation with an amplitude of 0.3A occurs, with angular velocity pulsation components. Synchronous generation of phase deviation, making Within 150μs, the positive value crossed zero and dropped to -2.4W, triggering the active damping injection module to generate a reverse current command with an amplitude of 0.8A. The measured assembly interface pressure converged from 12.5N to 5.2N.

[0033] To verify the synergistic effect and parameter boundaries of the technical solutions, this experiment set up a comparative system consisting of the sample group of this invention, a partially missing control group, and an out-of-range control group. Under severe collision conditions, the transient active power scalar reversal extreme value of the sample group of this invention reached -8.6W, and the system response time after the intervention of the active damping injection module was 110μs, limiting the maximum contact stress on the surface of the precision component to 15.6N. The partially missing control group removed the active damping injection module, retaining only the power polarity monitoring function, and the measured maximum contact stress reached 42.3N. Due to the lack of an active energy dissipation mechanism, stress concentration occurred in the component. The out-of-range control group reduced the cutoff frequency of the high-pass filter. Setting the frequency to 1200Hz and attenuating the power scalar amplitude by 65% ​​results in a 450μs delay in polarity reversal identification, causing the peak contact stress to rise to 31.8N. Based on the data evolution trend under different operating conditions, the sample group of this invention exhibits an injection intensity of reverse damping current when facing increased interference stress. The negative amplitude shows a monotonically increasing relationship. This control method based on energy flow polarity discrimination shifts the assembly overload risk from the non-real-time monitoring logic of the main controller to the physical execution layer of the servo drive unit. While avoiding the communication delay of the industrial bus, it uses the electromagnetic field of the controlled object itself to establish a dynamic barrier to block the rebound of mechanical potential energy, ensuring the structural integrity of precision components in the tight assembly process.

[0034] Example 3: This example combines Figures 1 to 2 The intelligent assembly monitoring and management system will be explained, such as... Figure 1 As shown, the system's control architecture includes a main controller and a servo drive unit. The main controller includes scheduling commands and alarm / interrupt functions, and is connected to the servo drive unit via a communication link marked with a real-time industrial bus. The servo drive unit contains, from top to bottom, a status acquisition module, a transient power flow analysis module, an active damping injection module, and a stator q-axis current control loop. The controlled object at the bottom of the architecture performs output current and position feedback operations, and transmits the current and position feedback values ​​upwards to the status acquisition module. This status acquisition module acquires the angular velocity feedback and outputs the angular velocity and current pulsation downwards to the transient power flow analysis module. This transient power flow analysis module extracts the pulsation to generate a power scalar. When the polarity of the active power scalar reverses, it triggers the active damping injection module downwards. This active damping injection module performs polarity reversal to generate a damping command and inputs the reverse damping current command downwards to the stator q-axis current control loop. This stator q-axis current control loop performs feedforward compensation to generate braking torque, ultimately dissipating mechanical rebound energy on the controlled object, completing the closed-loop control of the system's operating state.

[0035] like Figure 2 As shown, the overall hardware system consists of an end-effector, a servo drive actuator, and a host control node. The end-effector, located at the top level, encapsulates a servo motor electromagnetic actuator and position and current physical sensors. This end-effector is connected to the servo drive actuator via a field physical link to achieve current feedback and torque output. The servo drive actuator integrates five functional modules from top to bottom: a status acquisition module with a built-in high-frequency sampling clock (10kHz), a transient power flow analysis module with a built-in sliding shift register (100μs window), an active damping injection module with a preset negative power threshold judgment boundary, an impedance mapping module that performs transient impedance mapping as feedback gain, and an abnormal alarm module that performs real-time operating condition interference feature upload. The servo drive actuator is connected to the host control node via a real-time industrial bus to perform periodic polling and data synchronization. The host control node, located at the end level, deploys controlled task sequence scheduling logic and interference feature data processing unit.

[0036] Example 4: In an automated assembly line involving the precise mating of tenons and mortises for aero-engine blades, the assembly interface faces sub-micron level fit tolerances. Affected by fluctuations in lubricant film thickness, the mechanical system exhibits high-frequency nonlinear damping characteristics. When the tenon enters the interference zone, the transient contact stress rise rate exceeds 100 N / ms. Traditional monitoring architectures are limited by the 1ms to 10ms communication latency of industrial Ethernet, preventing the servo driver from shutting off torque before lattice crushing on the tenon surface. Therefore, the system needs to establish an in-situ energy suppression loop within the servo drive unit. The implementation logic of this technical solution under this condition is as follows: the system determines the sliding time window in the transient power flow analysis module. The value of this parameter is determined based on the mechanical resonance period of the assembly mechanism. Confirmed, satisfied Less than Physical constraints, in this embodiment The time interval is set to 100μs. Within a 10kHz current sampling interruption, the status acquisition module acquires the stator q-axis current feedback value and the rotor position feedback value, and calculates the instantaneous angular velocity of the rotor. A high-pass filter with a cutoff frequency of 500Hz filters out the low-frequency components in the above variables, generating the q-axis current pulsation component. With angular velocity pulsation components The multiply-accumulate operator inside the transient power flow analysis module is... Calculate transient active power scalar within a period The formula is as follows: ,in, For transient active power scalar, This refers to the pulsating component of the stator q-axis current. For the instantaneous angular velocity pulsation component of the rotor, when When the mechanical potential energy rebounds from the assembly end and is fed back to the motor shaft, the active damping injection module is triggered.

[0037] The implementation steps of the active damping injection module include obtaining the current control cycle. The absolute value of the negative value is compared with the preset damping gain coefficient. Multiply by each other to calculate the damping current compensation, where The value of is determined through offline calibration experiments and is used to achieve physical matching between the generated electromagnetic dissipation torque and the feedback interference potential energy. The calculated damping current compensation amount is used as a feedforward operator and is directly superimposed on the output register of the current regulator inside the servo drive unit, thereby changing the duty cycle of the PWM inverter. The injection of this damping current occurs after the original PI control logic operation of the servo system. Within a 100μs control cycle, a transient electromagnetic field opposite to the rotor motion direction is established, which converts the feedback mechanical potential energy into energy storage of the driver bus capacitor. The above mechanism establishes a microsecond-level energy closed loop at the physical execution layer, and sinks the process of suppressing assembly impact from the logic judgment of the main controller to the electromagnetic field response of the driver. This eliminates the masking of interference characteristics by the low-frequency load component in the mechanical transmission chain. The measured data shows that in the challenging scenario of 2μm axial misalignment during tenon and slot mating, the peak impact force at the assembly end is limited to less than 25N, avoiding surface scratches on precision components and providing deterministic physical safety protection at the system level.

[0038] Example 5: In the initial debugging scenario of a precision assembly station, the stiffness and moment of inertia of the mechanical transmission chain differ due to the influence of the physical mass of the end effector. The system determines the control parameters through a reference identification program. The state acquisition module, in a frequency sweep excitation state after the drive motor is locked at rest, generates a frequency sweep excitation signal in the range of 10Hz to 1000Hz with a step frequency of 1Hz, recording the amplitude-frequency characteristics of the stator q-axis current feedback value and the rotor angular velocity pulsation component. The transient power flow analysis module determines the mechanical resonant frequency based on the identified amplitude gain peak value. The system will set the cutoff frequency of the high-pass filter. Determined at At 0.5 harmonics, and calculate the sliding time window according to the following formula. The possible values ​​of: ,in, For sliding time windows, This is the mechanical resonant frequency; when the system experiences damping characteristic drift due to mechanical wear, the active damping injection module reconstructs the damping gain coefficient through an offline calibration process. The calibration program applies a step braking torque of known amplitude at the end of the simulated assembly, and simultaneously monitors the transient active power scalar at the moment of power polarity reversal. The system calculates the negative peak value and adjusts the damping gain coefficient in steps of 0.5 within the range of 0.1 to 10.0 until the injection of the current loop compensation parameter causes the attenuation rate of the angular velocity pulsation component to reach the critical damping state. The system then stores the corresponding value in the register and uses it as the current value. parameter.

[0039] Example 6: In the continuous operation of a precision assembly line, the current sensor generates zero-point offset noise due to temperature drift. This noise is superimposed on the actual feedback value of the stator q-axis current, causing a residual power scalar error in the system under interference-free conditions. To correct this physical deviation, the system executes a standardized zero-point calibration procedure before the start of each daily task. The status acquisition module collects 1024 cycles of current data stream while the drive motor is in a stationary locked state. The arithmetic mean of the current data stream is calculated to determine the current zero-point compensation amount. The transient power flow analysis module stores this value in a register and subtracts the current zero-point compensation from the real-time acquired actual feedback value of the stator q-axis current in subsequent assembly cycles. This is used to obtain the pulsating signal input after the bias component is removed.

[0040] When an industrial control cluster consisting of multiple six-degree-of-freedom assembly robots operates collaboratively, the system employs a synchronization procedure based on a distributed clock to eliminate command clock errors. The main controller calculates the global synchronization cycle based on the bus baud rate. The synchronization pulse is sent to each servo drive unit. The status acquisition module latches the actual feedback value of the stator q-axis current when it receives the rising edge of the synchronization pulse. The transient power flow analysis module calculates the transient active power scalar. Previously, the following formula was used to perform phase alignment compensation for the sampling time: ,in, For phase compensation angle, This refers to the instantaneous angular velocity pulsation component of the rotor. To measure the deviation time of the sampling trigger moment relative to the center point of the synchronization pulse, when the system is under heavy load assembly conditions, the multiply-accumulate unit synchronously extracts the second derivative of the stator q-axis current ripple component as an auxiliary characteristic quantity. The absolute value of the second derivative is less than zero and exceeds the current change rate threshold. At that time, the system confirmed that a substantial potential energy rebound had occurred at the assembly interface, exceeding the current change rate threshold. The system obtains the second-order derivative by statistically analyzing the current response curve of the controlled object under stall impact. Considering the impact of 20dB high-frequency electromagnetic noise in industrial environments on the calculation of higher-order derivatives, the system calls the data stream in the sliding shift register before extracting the second-order derivative, performs a 5-point least squares smoothing filter based on the Savitzky-Golay operator, and calculates the second-order rate of change using the central difference scheme on the smoothed current pulsation sequence. A second-order Butterworth low-pass filter with a cutoff frequency of 800Hz is used to further prune the differential result to remove pseudo-high-frequency components generated by numerical differentiation. This multi-stage filtering and low-order difference coupling strategy ensures that even under strong interference, the second-order derivative can accurately capture the current acceleration characteristics at the moment of assembly jamming. This multi-dimensional parameter coupling judgment method improves the system's adaptability to sensor drift and mechanical return gap pulses, enabling compensation commands to accurately apply to the actual collision suppression process. Based on the kinematics principle of multibody systems, the small displacement of the end effector in Cartesian space and the angular displacement of each joint space are determined by the Yarley method. A unique mapping is established using a comparable matrix. When the multi-axis integrated actuator detects that a specific motion axis meets the potential energy rebound judgment condition, the main controller synchronously extracts the real-time rotor position feedback values ​​of each motion axis in the bus network, calculates and updates the real-time Jacobian matrix corresponding to the current system structure, and calculates the pseudo-inverse matrix. The main controller then calculates the pseudo-inverse matrix, extracts the force vector in the direction of the motion axis that triggers the interference warning, multiplies it with the pseudo-inverse matrix, and solves for the joint compensation displacement vector in the orthogonal non-forced direction. After receiving the joint compensation displacement vector, the servo drive unit synchronously lowers the proportional gain parameter of the corresponding non-forced axis position control loop, inducing the mechanical actuator to generate the corresponding physical yield displacement along the non-forced direction. In this compliance protection mapping path, the system locks the components of each motion axis orthogonal to the interference force direction by calculating the pseudo-inverse matrix of the Jacobian matrix under the current pose. The calculated joint space compensation displacement vector is transformed into a scalar adjustment factor after dimensionality reduction through Euclidean norm operation and used as an independent variable input to a preset Sigmoid-type gain attenuation function. This function maps the magnitude change of the displacement vector to the proportional gain of the underlying position control loop. The reduced weights enable the actuator to instantly reduce the static stiffness of a specific motion axis when it senses interference potential energy. This allows it to dissipate collision stress by actively introducing flexible deformation in non-force-bearing directions, ultimately achieving closed-loop monitoring and proactive safety defense of precision component assembly status in complex production environments.

[0041] 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.

[0042] 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. An intelligent assembly monitoring and management system, characterized in that, Includes main controller and servo drive unit: The servo drive unit internally includes a mutually coupled state acquisition module, a transient power flow analysis module, and an active damping injection module; The status acquisition module is used to acquire the stator q-axis current feedback value and rotor position feedback value of the controlled object, and to perform differential calculation using the position deviation between adjacent sampling periods to obtain the instantaneous angular velocity of the rotor; The transient power flow analysis module is equipped with a high-pass filter with a cutoff frequency of 500Hz, which is used to strip the DC component from the stator q-axis current feedback value and the rotor instantaneous angular velocity to obtain the q-axis current pulsation component and the angular velocity pulsation component. The transient power flow analysis module is also equipped with a multiply-accumulate operator, which is used to discretely integrate the product of the q-axis current pulsation component and the angular velocity pulsation component within a sliding time window of 100μs to 500μs to generate an active power scalar that characterizes the direction of electromechanical energy conversion of the controlled object. The active damping injection module is connected to the transient power flow analysis module. It is used to read the positive and negative polarities of the active power scalar. When the active power scalar reverses from a positive value to a negative value, it determines that the controlled object has entered the mechanical energy feedback state, obtains the absolute value of the negative value, and proportionally maps it with the preset electromagnetic dissipation coefficient to generate a reverse damping current command. The servo drive unit is used to bypass the polling cycle of the main controller and inject the reverse damping current command as a feedforward compensation parameter into the stator q-axis current control loop of the servo drive unit to generate a braking torque opposite to the rotor rotation direction and dissipate the mechanical rebound energy generated by the controlled object.

2. The intelligent assembly monitoring and management system according to claim 1, characterized in that, The status acquisition module is equipped with a high-frequency sampling clock with a sampling frequency of not less than 10kHz. The status acquisition module uses the high-frequency sampling clock to sample the rotor position feedback value at equal intervals to obtain a discrete position sequence. The position deviation between two adjacent sampling periods in the discrete position sequence is calculated by the ratio of the position deviation to the reciprocal of the sampling frequency to determine the rotor instantaneous angular velocity. The rotor instantaneous angular velocity and the stator q-axis current feedback value are kept in phase synchronization on the time axis to eliminate the time-domain phase deviation when generating the active power scalar.

3. The intelligent assembly monitoring and management system according to claim 1, characterized in that, The high-pass filter is a second-order infinite impulse response filter; the transient power flow analysis module uses the high-pass filter to dynamically truncate the stator q-axis current feedback value, stripping away the low-frequency components that characterize the constant operating resistance of the controlled object, and locking in the pulsating characteristics that characterize the transient physical interference of the controlled object.

4. The intelligent assembly monitoring and management system according to claim 1, characterized in that, The transient power flow analysis module is equipped with a sliding shift register, which is used to store discrete data sequences within a sliding time window. Whenever the status acquisition module generates new sampling data, the sliding shift register removes the earliest set of data and pushes in the latest set of data, thus completing the real-time sliding update of the active power scalar.

5. The intelligent assembly monitoring and management system according to claim 1, characterized in that, The active damping injection module stores a preset negative power threshold. When the active power scalar is less than the negative power threshold, the active damping injection module determines that the controlled object has undergone a rigid collision and sends a reverse torque signal to the current loop controller of the servo drive unit.

6. The intelligent assembly monitoring and management system according to claim 1, characterized in that, The servo drive unit also has an abnormal alarm module, which is used to upload interference characteristic data that characterizes the real-time operating condition of the controlled object to the main controller at the same time as the active damping injection module generates the reverse damping current command, so that the main controller can interrupt the current controlled task sequence.

7. The intelligent assembly monitoring and management system according to claim 5, characterized in that, The negative power threshold is set based on the power fluctuation variance of the system under no-load conditions. The active damping injection module statistically samples the active power scalar under no-load conditions and sets three times the power fluctuation variance as the judgment boundary for the negative power threshold, which is used to shield the random electromagnetic noise inside the servo drive unit.

8. The intelligent assembly monitoring and management system according to claim 1, characterized in that, The servo drive unit is a multi-axis integrated driver. The status acquisition module simultaneously acquires the current feedback and position feedback of multiple motion axes. The transient power flow analysis module calculates the active power scalar corresponding to each motion axis. When the polarity of any motion axis is reversed, the active damping injection module locks all related controlled motion axes in a coordinated manner.

9. The intelligent assembly monitoring and management system according to claim 1, characterized in that, The servo drive unit also has an impedance mapping module; the impedance mapping module calculates the transient impedance characteristics of the controlled object based on the slope of the change of the active power scalar, and maps the transient impedance characteristics to the feedback gain of the active damping injection module, so as to realize compliant protection of the physical properties of different controlled objects.

Images